Skip to content
BY 4.0 license Open Access Published by De Gruyter January 31, 2023

Environmental landscape design and planning system based on computer vision and deep learning

  • Xiubo Chen EMAIL logo

Abstract

Environmental landscaping is known to build, plan, and manage landscapes that consider the ecology of a site and produce gardens that benefit both people and the rest of the ecosystem. Landscaping and the environment are combined in landscape design planning to provide holistic answers to complex issues. Seeding native species and eradicating alien species are just a few ways humans influence the region’s ecosystem. Landscape architecture is the design of landscapes, urban areas, or gardens and their modification. It comprises the construction of urban and rural landscapes via coordinating the creation and management of open spaces and economics, finding a job, and working within a confined project budget. There was a lot of discussion about global warming and water shortages. There is a lot of hope to be found even in the face of seemingly insurmountable obstacles. AI is becoming more significant in many urban landscape planning and design elements with the advent of web 4.0 and Human-Centred computing. It created a virtual reality-based landscape to create deep neural networks (DNNs) to make deep learning (DL) more user-friendly and efficient. Users may only manipulate physical items in this environment to manually construct neural networks. These setups are automatically converted into a model, and the real-time testing set is reported and aware of the DNN models that users are producing. This research presents a novel strategy for combining DL-DNN with landscape architecture, providing a long-term solution to the problem of environmental pollution. Carbon dioxide levels are constantly checked when green plants are in and around the house. Plants, on either hand, remove toxins from the air, making it easier to maintain a healthy environment. Human-centered Artificial Intelligence-based web 4.0 may be used to assess and evaluate the data model. The study findings can be sent back into the design process for further modification and optimization.

1 Discuss the use of computer vision and deep learning in an environmental landscape design and planning system

Environmental is the process of designing, creating, and maintaining landscapes while considering the area’s ecology to develop gardens that benefit both humans and the rest of the ecosystem [1]. Ecological landscaping requires a focus on conservation. Landscape gardening is more than just an art form that enhances the beauty of a landscape; it is functional and essential [2]. Plants that are an integral part of landscape design help reduce pollution and mitigate some of the effects of heat, sound, wind, and air [3]. People and other living things are increasingly influenced by landscape design to create conducive living conditions. Landscape design may significantly impact a community’s social, aesthetic, and environmental qualities [4]. They also consider the aesthetics of innovation to provide a natural atmosphere conducive to your health and the health of your loved ones [5].

Consequently, landscape design improves land management and reduces the environmental impact of urban living [6]. Landscaping improves air quality. Plants may purify both the air and soil. When it comes to purifying the air, trees are unbeatable. Pollutants such as smoke, ozone, and nitrogen oxides are absorbed by tree leaves, which act as a filter in the atmosphere [7]. There are both good and bad aspects to landscaping; it’s vital to understand understanding. As a result, it helps create an attractive and functional space to live, work, and enjoy one’s time [8,9]. Landscaping assists in the organization of areas to ensure long-term growth. Deforestation, pollution, and ecological change are only a few consequences. Furthermore, landscaping has several distinct environmental implications [10]. Landscaping will need to incorporate our knowledge of living systems into habitat design to reduce or eliminate the use of natural resources while still providing places of cultural and aesthetic significance for the people who live in them in the future [11,12].

The transmitter and receiver layers of a deep neural network (DNN) are separated by many layers. Neurons, synapses, weights, biases, and functions are the same in all neural networks, regardless of the kind [13]. One of the critical advantages of deep learning is that identifying hidden patterns in the data or a profound grasp of complicated interactions between interdependent variables is one of the primary benefits [14]. Computer vision, voice identification, natural language processing, machine translation, and bioinformatics have been gained from deep learning architectures like DNN, deep belief networks, and recurrent neural networks [15]. In its most basic form, a DNN, short for “deep net,” is a neural network with at least two layers of complexity [16]. Deep nets use advanced mathematical models to analyze data more complicatedly [17]. DNN is machine learning in which high-level functions are derived from input data using many layers of nodes [18]. To identify the item, creative and analytical components of information are examined and organized. According to popular belief, DNN designs are superior in generality and reduced regularization [19]. For instance, face identification develops hierarchies of representations: edges and lines at the bottom, eyes close to the top, and whole faces at the top. A DNN is useful when replacing human labor with autonomous work that does not compromise effectiveness [20]. DNN may be used in a variety of ways in the real world. Deep learning designs take simple neural networks to the next level [21]. Deep learning networks may be built by data scientists using these layers to enable machine learning, which can train a computer to correctly mimic human skills, such as recognizing speech, classifying photos, or generating predictions [22]. When it comes to deep learning, the primary benefit is that it tries to learn high-level characteristics incrementally from data [23]. As a result, domain expertise and complex core feature extraction are no longer required. From autonomous driving to medical gadgets, deep learning applications may be everywhere [24]. Automated Driving: Automotive researchers use deep learning to recognize stop signs and traffic signals. Pedestrians may also be detected using deep understanding, which reduces the number of collisions.

This manuscript investigates human-centered artificial intelligence (AI) and Web 4.0 technologies in current landscape design and topological improvement applications. With the fast growth of human-centered AI technology, basic machine logic has been unable to keep up with people’s demands for the intelligence of different electronic gadgets. As a result, human-centered AI has developed at a critical juncture in history to satisfy these demands. As a result, two innovations are offered, both inspired by the growing growth of human-centered AI. (1) Using the core perspective of data mining, human-centered AI investigates the new visual phenomenon occurring in the context of Web 4.0, as well as cross-border and multidimensional interrelations characterized by visual images. (2) After correction, Web 4.0 demonstrates that the same metric information can be found between the prediction exterior and the correlating image pore structure as demonstrated in a three-dimensional (3D) scenario.

The main objectives of the article are as follows:

  • Sustainable or environmental is a way to save money and effort by designing, creating, and ecologically maintaining your landscape. Environmentally compatible landscape design provides wildlife habitat while reducing pollutant emissions and creating healthy recreational areas.

  • As landscape architects, humans must understand living systems to develop habitats that help minimize the consumption of natural resources while providing places of value and beauty for people.

  • The design of the landscape environment, society, culture, and social interactions are intertwined in research, and the game’s rules demand it. To undertake basic and applied research, we must work together across disciplines.

  • Proposed DL-DNN can learn and create output that is not constrained by the input data they receive. Data loss does not affect if it functions because it does not rely on a central database to record its input.

  • Real-world scene models and their matching rates have been enhanced in simulations, allowing for the integration of 3D images based on human-centered AI technologies to be realized.

The rest of the article follows Section 2 for the literature survey of the existing method, Section 3 for proposed method for DL-DNN to be discussed, Section 4 for experimental analysis, and Section 5 is the conclusion of the paper.

2 Concept of environmental landscape design and planning system

Fan et al. [25] described the 3D Mesh Simplification Algorithm for Urban Landscape Ecological Design. Urban road environment environmental development is characterized by the ecological theory, which may satisfy the regular traffic function of the road and create a nice green landscape. This research provides a 3D network reconstruction approach and vision technology and proposes an urban landscape ecological design visualization scheme. Using model design and simulations, the suggested framework’s performance is confirmed. Compared to current models, the suggested one exceeds them all.

Araújo et al. [26] deliberated the Agriculture 4.0 Landscape. If this fourth revolution is successful, it might help improve agricultural development, providing the demands of the world population in a fair, resilient, and sustainable manner. As a final step, we describe a high-level cloud-based IoT architecture that may be used to build future smart agricultural systems. An understanding of the idea of Agriculture 4.0 and recommendations for its effective implementation is anticipated to have a beneficial influence on the research in this area effort is expected to have a favorable impact on that research.

O’Neill and Maravelias [27] discussed that instead of looking at challenges in isolation, researchers might be able to uncover more efficient farm-to-fuel production systems if they combine the landscape design and biomass supply chain network design challenges. Improved landscape-specific uncertainty may now be accounted for due to the SCND’s extension. In contrast, feed supply can be better controlled, and greenhouse gas emissions may be predicted more accurately. Recent developments in field-scale crop yield, land-use change modeling, and upgraded SCND models focused on exact spatial resolutions facilitated the creation of integrated models. Despite early attempts, the computational burden of spatially explicit models makes it challenging to integrate landscape design with SCND.

Kang [28] discussed infrastructural support for the geographic data-enhanced understanding of simulations and the ability to see both natural and manufactured landscapes, as well as the widespread acceptance of the Wireless Internet of Things (WIoT). The (WIoT) enables everyday objects to become smarter, more capable, and more communicative than ever before. Ultimately, the WIoT wants its landscape design, and this development opens up Architecture-specific research and the normal architectural landscape area. The system briefly covers landscape applications for street architecture and advocates a street landscape design environment for WIoT system architecture. In the proposed system’s description, intelligent design, street architecture, and wireless network connectivity in the landscape are all described as things, whereas mapping is described as network technology.

Xie [29] introduced a wave of information that has rushed across the globe due to the fast growth of information and communication technology (ICT). Now that ICT has advanced, many people believe it is an essential component of urban competitiveness and a fundamental skill for today’s workforce. People’s lives and work methods have altered due to the rapid development of ICT. Because of this, the city’s general landscape planning and design, as well as competitive assessment to establish suitable solutions, are critical to its long-term growth.

Zhang [30] proposed that virtual reality technology (VRT) provides immersive vision and intuitive interaction. Any biomedical professional may utilize the VRT to create an image classification model with its help, build students’ capacity for independent research, and transmit knowledge and innovation. Deep learning and the early stages of comprehension can benefit from the three qualities of VRT. Model creation was also brought up during the investigation, feedback, and assessment phases. As a result, students get greater control over their education through a more in-depth comprehension of the work of academics involved in citation-related studies. This is the basis for VRT-based deep learning and a link between the baseline model and the model for providing feedback and evaluating learning results.

Badach and Raszeja [31] introduced the need to provide a conceptual framework for using landscape and green space indicators (LGI) in urban planning that incorporates ecological, structural, and aesthetic factors. Consequently, it is possible that the adoption of the proposed LGI framework would considerably improve the environmental and aesthetic quality and structural variety of the urban landscape and green space in the urban environment. According to the LGI framework proposed in this article, planners, policymakers, and other stakeholders can use the framework to improve the governance of urban landscapes and urban green space management and better account for human health and well-being when deciding how to manage urban green space.

Liao et al. [32] introduced the development of epoch as a period in which people and their activities affect nearly every aspect of the planet’s ecosystems. Landscaping sustainability science (LSS) is an attempt to understand better and strengthen the dynamic relationship between ecosystem services (ES) and human well-being in the face of increasing uncertainty. Using LSS as a lens, we explore the theoretical underpinnings of the field, the most recent advances in research methods, and the potential for LSS to be applied across diverse landscapes to address sustainability concerns.

Fan et al. [33] detailed the progress in urban road building and landscape design co-occurring in China. As conduct, the aesthetic design does not rely on science’s subjective nature. It is only in this fashion that works of art that fit a particular group’s aesthetic ideals may be developed. The ecological theory guides the eco-environmental design of urban roads, which can satisfy the regular traffic function of the road while also creating a lovely green environment. This research presents an urban landscape ecological design technique using three-dimensional network reconstruction and vision technologies.

Li and Xu [34] discussed that simulating the planning effect is extremely important for ensuring the rationale of landscape design. Technology that uses three-dimensional images to overcome the flaws of standard landscape distribution modeling technology is created in this research. First and foremost, an image-based three-dimensional analysis approach is provided to distribute landscape gardens. Moreover, the case study and performance test results show that the achieved distribution rationality technology is better than the competition. This technique may be a scientific reference and basis for adopting three-dimensional photographs and other new technology for the current landscape gardening sector.

Brown et al. [35] introduced landscape design and planning. This study examines the utility of social media in identifying public opinions of intrinsic landscape values. To guide future landscape design and planning, they look at individuals’ and institutions’ values and publicly debate landscape views and challenges and how they could be improved. They further look at the location, substance, and mood of tweets. Our goal is to devise a mechanism for landscape architects and planners to use social media data to gauge public opinion on various aspects of landscape design. However, our initial study provides clear instructions on effectively acquiring and interpreting this data for landscape planning and design.

Picchi et al. [36] proposed landscapes worldwide are beginning to change due to the transition to a low-carbon future. Renewable energy solutions must not compromise other ecosystem functions, such as food production, allowing long-term landscape shifts. The second issue is the lack of practical methodologies and spatial reference systems that consider cultural and regulatory ecosystem services in applications that address ecosystem services for landscape planning and design. The application of participatory mapping and landscape modeling study for cultural ecosystem services and landscape design principles should be the focus of future research.

Alpak et al. [37] detailed that landscape architects develop the environment according to the varied roles of buildings, which are an organic part and an outside extension. Creating a cohesive whole by merging various elements is one of the most important design goals. A distinctive design product can only be achieved by establishing connections between the many components of the organization as a whole. This way, students will learn how to create well-used open spaces, one of today’s most pressing issues. After that, a survey was done to see how the systems approach affected the student’s ability to develop innovative and valuable ideas. This study shows the need for a systems perspective for design education and urban design.

Seçkin [38] introduced an effective building design that relies heavily on landscape design. It is possible to shelter buildings from the sun, protect them from strong winds, and create possibilities for natural ventilation by incorporating landscape aspects into a structure. Among other things, landscape components may assist in cleaning the air and water, absorbing floods, improving aesthetics, providing recreational amenities, and providing biological homes for species. For the most part, it makes sense to plant trees on a structure’s east and west sides, windbreaks on the north, and open fields on the south. Some of the most important principles and subjects in landscape design about sustainable building design are covered in the following sections.

3 Design and planning system for environmentally sustainable landscapes

As users consider the environmental effect of the design, it is essential to include components that are both attractive and useful to the natural world. Wildflowers and shrubs were substituted for the lawn, which the homeowners preferred to do away with. It is possible to sequester carbon in landscapes while still cleaning the air and water and increasing energy efficiency. These landscapes ultimately produce value because of the positive effects on the economy, society, and the environment. The afforestation initiative, which involves planting trees along the edge of a desert to counteract desertification and land degradation, is the green belt in these regions. Improved quality of life and natural resource preservation are all sustainability benefits. Going green and sustainable is good for its bottom line but also for the environment.

Digital landscape architecture planning and design are in-depth detailed using the supervision of a human-centered AI-based web 4.0 as the primary body of the original article. The system of digital landscape design and construction was built using special models. Quantifiable materials gathered through research and information collecting are translated into variables and parameters by software calculation and analysis algorithms to precisely characterize various environmental conditions and accurately represent the location. Design components are linked to site circumstances using a particular design model depending on each design’s kind and target audience. This article examines the use of human-centered AI-based web 4.0 in landscape design and topology optimization. In product semantics, the visual properties of manufactured goods are examined under various usage conditions. It is possible to achieve semantic communication between people and objects by incorporating design product semantics into creating new products. Product semantics primarily focuses on how people and things communicate and communicate to consumers the product’s content and its visual features, such as shape and color.

Human-centered AI-based web 4.0 is a low-stress environment with few distractions. The human-centered AI-based web 4.0 item may be gripped, pitched, and positioned by the user. A human-centered AI-based web 4.0 user interface was used with various tools for a more comprehensive understanding of the model and its environment. Figure 1 shows multiple elements that influence the environment, including climatic ones like rainfall, temperature, and wind speed; reproductive ones like predators and parasites; soil microorganisms; and vegetation like pH, texture, and structure. Concentration describes the volume of gas in the air when measuring gases like carbon dioxide, oxygen, or methane. Whether it is raining on land or water, the amount of rain that falls on Earth’s surface is rainfall. It forms when air masses pass over warm bodies of water or damp ground. This water is finally released by the clouds and falls as rain. Understanding how ocean circulation patterns affect global climate and regional marine ecosystems, both naturally occurring and caused by humans is critical. Several extreme kinds of weather and climate occurrences fall under this category: unexpected, uncommon, severe, or unusual weather. According to the theory of “social motivation,” humans are driven by their need to connect with everyone else. People don’t do their own, which has been critical to human survival throughout history. Biodiversity in cities refers to the wide range of species and biological systems. As a whole, urban biodiversity is affected by various causes, including both natural and artificial ones. A person’s capacity to live a life they value is defined as human well-being, which encompasses access to land and natural resources and the ability to earn a living.

(1) W = ( X ) Y X × π 2 ÷ ( r ) .

Equation (1) denotes W for the development of environmental change, X for landscapes, Y for rainfall, and π is the mathematical function for temperature. Both natural and human-caused factors contribute to environmental change. Large amounts of energy and materials are transformed and transported by ecological systems and human activities, contributing to environmental changes.

(2) P = ( U + V ) U V × max 2 V .

Equation (2) denotes P for the impact of rainfall, V for temperature, max 2 for the maximum in landscape design. Due to precipitation patterns, people and ecosystems will be affected by changes in water availability throughout the year. Floods and droughts are expected to become more frequent and severe due to the changing precipitation patterns. They should expect an increase in the intensity and frequency of wildfires in the following years.

(3) Q = R S tan θ ± R * .

Equation (3) says Q for human well-being in the environmental, R for concentrations, S for biodiversity, tan is the trigonometric function for landscaping decisions, and θ is the mathematical function for extreme events. A wide range of factors contribute to a person’s sense of well-being. It includes financial stability, family connections, and physical and emotional well-being. It gives a place to work and play, a sense of community, and a sense of personal safety.

Figure 1 
               Environmental landscape.
Figure 1

Environmental landscape.

Figure 2 shows the digital landscape. The planning and design process of human-centred AI-based Web 4.0 architecture is explained theoretically and technologically using the paradigm and model. Designers use their experiences and sentiments to guide them through the conventional design process. They begin by thinking about a project’s overall strategy, then use 3D design software modeling to sketch out a rough map and tweak and modify the plan as they go along. There are various steps to this design process. Regarding this issue, the planning process is unclear. It is impossible to accurately represent a whole location due to a strong desire to express the designer’s preferences, and design outputs are subjective.

Figure 2 
               Digital landscape.
Figure 2

Digital landscape.

Using the digital model, the designer can establish parameters that influence design, link the various design system components, and finally govern the development of a final design result through computer-based parameters. With this process logic approach, designers may better account for the site’s actual circumstances and how different system parts contribute to the final design. Landscape design requires qualitative and quantitative study, and modeling is integral to this process. When combining the two, it is possible to have two sorts of relationships between qualitative and quantitative research. First, the temporal connection relates to the order in which both occurred in the research study. The second is the link between the main and secondary methodologies utilized in the study. There are three distinct types of mixed methods: sequential, where qualitative and quantitative research are combined in the same study, as well as the simultaneous use of both styles; primary and secondary, in which qualitative and quantitative data are combined; and comprehensive, which combines all three types of mixed methods. Qualitative and quantitative methodologies are intertwined in the digital landscape planning and design paradigm. The model uses qualitative and quantitative methodologies simultaneously, complementing each other.

The following are the stages of a landscape design algorithm:

Stage 1: Initialize the input data.

Stage 2: Analysis is done by analyzing and comparing pre-processed picture data sets.

(4) U j = l = 1 M ( V jl ) n Y l l = 1 M ( V jl ) n .

As shown in equation (4) input cluster center has been calculated. U j is an input cluster center, and V jl is a value of the grayscale.

Stage 3: Identify the picture values and the surrounding values. Deep neural networks are used to both train and verify the values.

(5) V jl 1 0 c jl ( s ) = min 1 i m d { c jl ( s ) } .

As found in equation (5), data partition degree value has been deliberated. c jl ( s ) is the data partition degree value.

Stage 4: Make sure that the convergence value is in good condition.

(6) µ > Vs 1 Vs .

Stage 5: Analyze the picture and assign a value to each landscape’s environmental attributes.

Stage 6: End.

Figure 3 shows that the particular challenge strategy that relies on practical methods or shortcuts to provide answers that may not be optimum but are sufficient in the face of a restricted time or deadline is called a heuristic or heuristic methodology. Landscape design and planning are primarily concerned with whether or not a proposed plan can coexist peacefully with the surrounding environment and whether or not the proposed vertical design is closely integrated with the site’s existing topography. The human-centered AI-based web 4.0 software plug-in will unite the coordinates. The planning arrangement and the status quo scheme can be displayed in the base layer, and the roll-up function can be used to compare various strategies. Products and features are defined by their value and purpose in product specifications. It is created by the product manager to explain what you’re developing, who you’re building it for, and how it helps the user. Multiple dependent variables lead to a single outcome when modeling is multivariate. This helps to explain why the vast majority of real-world problems are multidimensional. For example, humans cannot predict the weather for any given year based on the season alone. Numerous variables, such as pollution, humidity, and precipitation, play a role. As distinct from construction-related expertise, architecture is designing and constructing structures. As a result, architecture is used to meet both practical and expressive needs, and as a result, it has a dual purpose. There are several phases in a biological organism’s life cycle. All organisms, including plants and animals, have a life cycle. There are several ways in which diagrams may illustrate the many stages of a plant’s life cycle. Repeating life cycles are commonplace. A multidimensional spatial surface represents a population’s average fitness in an adaptive landscape. People are shown as a single dot on the screen, whereas an entire cloud of dot images represents groups. One defines probability distributions for the model’s parameters before viewing data to incorporate prior knowledge into Bayesian probabilistic modeling. Based on the evidence, these previous views are re-evaluated as posterior beliefs. Evolutionary biologists use fitness or adaptable landscapes to depict the link between genetics and reproductive success. Every race is believed to have a specific replication rate.

(7) M = u v y * min 1 u .

Equation (7) denotes M for the evolution of the adaptive landscape, is the mathematical function for product requirement, y is the muti-variate, u is the lifecycle, v is the heuristics inputs, and min 1 is the minimum optimization. Evolutionary changes that make an organism more suited to its environment are called “adaptive evolution.” The alterations lead to a better likelihood of surviving and reproducing, and the modifications make it possible for a particular organism to adapt to a new environment.

(8) O = q × z q + 2 log ( Z ) .

The equation (8) says O for the necessary Bayesian probabilistic. q for design space, Z for the adaptive, and log is the logarithmic function for the product requirement. According to the Bayesian probabilistic model, a hypothesis’s probability can be evaluated by specifying a prior chance. As new information is gathered, it is modified to reflect this further information and produce a new posterior probability.

(9) N = o + p p o sin 1 ( * ) .

Figure 3 
               Landscape planning system.
Figure 3

Landscape planning system.

Equation (9) denotes N for the environmental life cycle. sin 1 is the trigonometric function for architecture, is the mathematical function for adaptive landscape, p for probability, and o variant plan. Resources are used throughout a product’s life, and the environment is somehow impacted. To lessen the quantity of energy used in the production process, it may be necessary to utilize more material or use more power at the product’s disposal stage.

Figure 4 focuses on a property’s overall master landscape planning and the detailed garden design of landscape elements and plants inside it. The landscape design incorporates features such as functionality, aesthetics, horticulture, and environmental sustainability, often broken down into various phases of the design process. Human-centered AI-based web 4.0 technology to design landscapes can effectively save a lot of workforce, reducing costs; the developer can modify or modify the plan to their own needs and relatively close to the precise requirements of its users for landscape effect to obtain excellent interaction between user and design. Human-centered AI-based web 4.0 will considerably cut landscape design time, make the design more sensitive to actual user demands, enhance design entirely, and make it simpler for designers to reconfigure the design. Landscape designers often collaborate with architects, civil engineers, surveyors, contractors, and craft talents to create beautiful landscapes. Increasing one’s productivity is a common goal since many people believe that doing so will save time. They’ve gained time but still don’t have additional time, so the conventional approach is to occupy that time with other activities. Improved efficiency was only feasible for a certain amount, which means higher efficiency, better techniques, better management, and decreased overheads.

Figure 4 
               Landscape design.
Figure 4

Landscape design.

The purpose is to gain higher efficiency and better value for money. Analyzing various options’ relative costs and benefits is known as cost-effectiveness analysis. There is a distinct difference between a cost-effectiveness analysis and a cost–benefit analysis. Increased productivity implies more output is generated from the same number of inputs. Production functions are employed to measure a particular system’s productivity by providing relevant information. One of the best ways to increase a company’s profitability is to reduce waste, and waste elimination is a key element of lean manufacturing. Any product, instrument, or process that is ineffective or does not provide value is considered a waste. The accuracy of data, measurements, and statistics may be verified down to the smallest of details.

(10) H = ( n ) × log ( L ) ÷ L n .

The equation (10) denotes H for landscape evolution, n for saving time, the log is the logarithmic function for efficiency, and L for accuracy. Landscape evaluation finds them particularly sensitive and prized for their unique characteristics and aesthetic appeal regarding landscapes. Protected zones can be designated for these terrains, and local and national landscape designations affect local identities.

(11) J = ( ω ) s ˆ × cot ( π | 2 ) .

Equation (11) says J for critical of enhanced productivity, cot is the mathematical function for wastage, π is the mathematical function for resources, and s for effective. As a result, the workplace culture and morale improve due to increased productivity. In most cases, if a company is highly productive, it will eventually be profitable, and as a result, employees will be given incentives.

(12) F = π 2 + a + b sin ( β ) .

Equation (12) denotes F for landscape efficiency, π is the mathematical function for eliminating well-being, sin is the trigonometric function for productivity, b for cost-effective, and β is the mathematical function. An energy-efficient landscaping strategy might include employing locally produced materials, onsite composting and chipping to cut down on green waste hauling, hand tools rather than gasoline-powered equipment, drought-resistant plants in arid areas, and stock from local growers to cut down on the amount of energy used in transportation.

Figure 5 shows the information delivered to a computer for processing, known as input. A keyboard, mouse, or other input device is commonly used to provide data to the computer. A 3D database and virtual scene model of the landscape are created using human-centered AI-based web 4.0 to create a 3D visual environment planning and design. A 3D visual overall urban modeling is performed to provide a three-dimensional representation of the virtual scene landscape. Lastly, a computer simulation experiment has been carried out, and the analysis of its results to accurately determine the location of the actual model and the matching rate, to integrate 3D image, 3D stereo, and unified communications technologies, and to accomplish the landscape planning assignment. The restrictions of matrix factorization can be overcome by using deep neural network (DNN) models. Due to the adaptability of the network’s input layer, DNNs may readily add query features and item features, which can better capture a user’s unique interests and increase the accuracy of their suggestions.

Figure 5 
               Landscapes in deep learning.
Figure 5

Landscapes in deep learning.

To put it another way, the process of putting data into a computer is known as input. To train a deep learning model, vast data sets of label data are used, and neural network designs learn features directly from the information. One of the most prevalent types of deep neural networks is deep neural networks. Natural language processing, speech recognition, and other applications frequently use recurrent neural networks. Recurrent neural networks analyze data’s sequential properties to predict the following situation. A more explainable model is more likely to be generally applicable. A model’s validity does not need to comprehend all of the model’s details for every data point. A better solution requires good data, a model, and knowledge of the problem. It is easy to interpret when people can quickly grasp the logic behind the model’s predictions and judgments. For a model to be understood and trusted, it must be subject to interpretation. All aspects of daily life involve probability, from sports to weather reports to blood tests to guessing a baby’s gender in the womb to congenital impairments to statics.

Probability is the subject of this chapter. When a firm engages directly with its target audience, it is referred to as brand activation. Through various events or marketing efforts, as long as the action is focused on a specific customer experience that helps create a company’s reputation, this may be done successfully. Grouping things together based on a shared characteristic is known as classification. Decision-making information should be presented in the same way as a narrative. The message, the audience, and the message-to-audience match must be all clear before you begin. On the other hand, data presentations typically fall short of these simple guidelines.

(13) E = cot ( α + β ) ÷ ( j 1 ) h .

Equation (13) denotes E for used for deep learning; the cot is the trigonometric function for input data, α is the mathematical function for network, β for mathematical function in model, j for activation, and h for probability. Deep learning is applied in sectors ranging from self-driving cars to medical gadgets. Researchers in automated driving are utilizing deep learning to identify things such as traffic signals and stop signs. Pedestrians are also detected using deep learning, reducing the likelihood of an accident.

(14) G = ( I + 1 ) g × max 2 ( g ) .

Equation (14) says G for recurrent network analysis, I for the interpretable model, max 2 for maximum for providing information, and g for decision making. Neurons in recurrent networks have both feedback and feed-forward connections. Simply described, it is a neural network with loops linking the output responses to the underlying neural network input layer. As a result, the network’s output replies serve as extra input variables.

(15) D = l M ± 1 2 ( l × M ) .

The equation (15) denotes D for the decision-making environment, l for information, M for feature importance. The setting in which decisions are made It is claimed that the significant role of every manager is to make decisions. In addition to making intelligent judgments in the company’s best interest, managers undergo a series of sequential procedures. The term decision-making process refers to this procedure.

Figure 6 shows the term environment; it refers to the entirety of our planet’s physical surroundings. Human-centered AI on the web 4.0 is combined with a DNN to create a 3D database and virtual scene model of the landscape for 3D visual landscape planning and design and 3D visual urban landscape modeling to produce a 3D representation of the virtual scene. Everything in and around us, whether alive or not, is considered part of the environment. Nonliving elements of the environment are essential to the survival of all living creatures. Command, restrict, or manage are synonyms for the verb “control.” Keeping an eye on the overall coordination is one way to exercise control. Soil and air are both cleaned by plants. Tree leaves absorb and remove pollutants from the air, including smoking, ozone, and nitrogen oxides. Adding additional greenery to a city’s landscape can help to reduce air pollution.

Figure 6 
               Environmental of deep neural network.
Figure 6

Environmental of deep neural network.

On the other hand, a company’s application landscape is a collection of related programs. One way to think about it is to see it as a visual representation of an application’s whole landscape. As with making sense of the environment with our senses, computers are capable of “machine perception,” a comparable skill. In a controller, like the neurons in a real brain, the nodes in a neural network are organized in layers. Every layer has its network, and all of the nodes are linked to each other. As the number of layers increases, people say that the network has more depth. Millions of signals can be sent to a single neuron in the human brain. Attached hardware is the primary way computers sense and react to their surroundings. The net input in the first layer of the network is equal to the information multiplied by the weight and the bias. A modest weight is required for high intakes to keep the transfer function from saturated. It generates a probability distribution in which the outputs add up to one – zero through one of the possible outcome values for each class. One is the total of the probabilities of each type. The course with the highest probability output is the one predicted.

(16) A = ( x ) d x + cos ( t ) π 2 .

Equation (16) denotes A for a deep neural network in control, x for input, d for output, t for the environment, cos is the mathematical function for the plant, and π is the mathematical function for decision. By adjusting the weights and biases of a neural network, it is possible to design several different processing systems based on learning algorithms. Activation functions and alternative learning techniques are discussed as well.

(17) B = tan ( θ ) c + 1 e × log ( π ) .

Equation (17) denotes B for an end-to-end control deep neural network, θ is the mathematical function for the environment, tan is the trigonometric function for the neural program, c for application, the log is the logarithmic function for processing, e for perception, and π is the mathematical function for controlled. To train a potentially complicated learning system, end-to-end refers to teaching it using a single model representing the whole target system without the intermediary layers typically seen in standard pipeline designs.

(18) C = ( α + β ) i max 1 ( k ) i k .

Equation (18) denotes C for detection of plant, α is the mathematical function for a loan, max 1 for a maximum of programming, k for a loan application, β is the mathematical function for target, and i for the sorted array. In machine vision, the identification of plant diseases and pests is critical. Machine vision equipment is used to capture photographs, which are then analyzed to determine whether or not obtained plant images include disease or problems.

Figure 7 shows that when it comes to spatial planning, we’re discussing making plans to affect social, economic, and environmental change and creating visual representations of where these activities should occur. To put it another way, regional planning is concerned with land-use activities, infrastructure, and settlement expansion distributed throughout a broader land region. Laws governing these regions’ effective development and management are also included. To maintain sustained human settlements, land use planning allocates land between competing and often conflicting purposes rationally and environmentally soundly. Intellectual development and management of our landscape depend on proper land use planning. Residential, agricultural, recreational, transit, and commercial examples of land use. The land is a professional-grade landscape design and irrigation built open. Users may focus on the art of the design process with total precision, flexibility, and precise control of inland planting design. This final landscape design includes all of the components of the landscape.

Figure 7 
               Environment of the landscape planning system.
Figure 7

Environment of the landscape planning system.

In most cases, this takes the form of a hand-drawn illustration on effect. They can learn much about the space we have to work with from an open space layout. Developing objectives and selecting open space parcels based on purpose, distinctiveness, value, and availability are critical components of a plan (Agency for Housing and Community Renewal (HAD)). Examining how something will affect the environment (EIA). In the short term, environmental impact assessments (EIAs) are used to determine potential environmental impacts and hazards associated with proposed development projects’ rate of infant mortality (IMR).

(19) Q = a cot ( a + b ) × b a + c .

Equation (19) denotes Q for landscape performance; the cot is the mathematical function for the plan, a for open space, c for land use plan, and b for assessment. To put it another way, landscape performance refers to how well a landscape solution accomplishes its stated goal while promoting long-term viability. A landscape’s performance may be measured by its ability to achieve environmental, social, and economic objectives.

(20) R = 1 s u u s * log ( w ) .

Equation (20) denotes R for developing the regional plan . The log is the logarithmic function for the structure plan, u for the binding land plan, w for the spatial project, and s for the program. While national policies and concerns are covered more thoroughly in regional plans, they also help to improve integrated development at the regional level. A city’s planning area appears small in comparison to a region. Rural and underdeveloped areas are included in the regional plan, which is not often included in local programs.

(21) S = ( U + 1 ) ÷ T min 2 x .

Equation (21) denotes S for spatial planning development, min 2 for maximum landscape programs, T for SEA, U for open space, and x for planning. To enhance a more equitable distribution of economic growth within a particular region than market forces would otherwise generate, spatial planning includes efforts to coordinate and improve the geographical implications of other policy areas.

4 Experimental analysis of environmental landscape design

Environmental analysis refers to examining all of the factors, both internal and external, which affect an organization’s success. It determines whether or not the organization’s stated goals can be met with the current strategy. A site study helps identify and address the property’s particular issues to build a functional and comfortable outdoor living area. Additionally, the site study should incorporate critical aspects of landscape design that should be considered. It is a method in human geography that studies and describes a landscape to determine how humans have affected it. This study is used to put human-landscape interaction into context and aid a plan for the future usage or repair of the landscape.

4.1 Analysis of landscape-level index

Figure 8 shows that given nature, a broad definition of approach promotes fair and sustainable land use while enhancing efforts to reduce and adapt to climate change within a given region through integrating theory and policy for various land uses. It is a broad definition. Ecological field landscape ecology focuses on the patterns and interactions between groups that make up vast regions. On a landscape at this scale, the distribution of soil types can have significant ecological implications when compared to the other existing methods such as Wireless Internet of Things (WIoT), supply chain network design (SCND), information and communication technology (ICT), virtual reality technology (VRT), and landscape and green space indicators (LGI).

Figure 8 
                  Landscape-level index.
Figure 8

Landscape-level index.

4.2 Distribution of landscape

Figure 9 says landscape design concepts include line, shape, texture, and colors to create harmony, unity, scale, variation, emphasis, and sequence. All these things are linked together – proportion, order, repetition, and unity guide landscape design ideas. Objects in a landscape are referred to as being in proportion to each other by their relative size, and plants and landscapes should be balanced. The organization and harmony of a landscape are two examples of what we mean by the concept of control.

Figure 9 
                  Distribution of landscape.
Figure 9

Distribution of landscape.

4.3 The rationality of landscape

Figure 10 says mass, shape, line, texture, and colors are some design components that fall under this category. They are employed throughout the landscape to modify the environment and provide visitors with a one-of-a-kind experience. Landscaping and landscape architecture are all terms used to describe the process of organizing and changing existing characteristics in a natural or built environment. Urban and rural landscapes may be created by planning, designing, and managing open areas.

Figure 10 
                  Rationality of landscape.
Figure 10

Rationality of landscape.

4.4 Administration of land-use changes

Figure 11 shows human activities that alter the physical properties of the land through emissions of heat-trapping greenhouse gases due to changes in land use and cover. Getting approval from the appropriate tax body is the first stage in land conversion. It is necessary to submit a “change of land use” application to the relevant authorities, explaining the transformation’s objective.

Figure 11 
                  Administration of land-use changes.
Figure 11

Administration of land-use changes.

4.5 Effect of the environment

Figure 12 shows that air, water, soil, noise, and light pollution too are causes of natural effects. This is called environmental pollution when Earth’s physical and biological components are polluted to such an extent that regular environmental processes are negatively affected. People who are exposed to green, natural areas have better mental health. Stress reduction, an increase in good emotions, cognitive restoration, and beneficial effects on self-regulation are some of the mechanisms that may be involved.

Figure 12 
                  Effect of environmental.
Figure 12

Effect of environmental.

4.6 Characteristics of land use carbon emission

Figure 13 shows that the severity of global warming is exacerbated by how we use land. Many greenhouse gas emissions come from land use, including agriculture, forestry, and other activities. Changes in land use, such as removing trees to make way for farming, contribute to emissions. Fossil fuel combustion for power, heat, and transportation is the primary source of greenhouse gas emissions in the United States. The transportation sector emits many of the world’s greenhouse gas emissions.

Figure 13 
                  Characteristics of land use carbon emission.
Figure 13

Characteristics of land use carbon emission.

4.7 Judgment rate (%)

In the field, data have been gathered and analyzed as indicated in Table 1, which compares the results of the standard technique with those from this study. During the 0–10 range, the conventional approach’s judgment rate steadily rises, although it is always lower than the 3D simulation method.

Table 1

Judgment rate (%)

Time WIoT ICT VRI LGI DL-DNN
5 67.91 45.34 65.09 33.34 76.03
10 64.22 52.05 63.03 43.14 74.06
15 60.45 56.31 66.21 36.19 78.43
20 54.51 59.46 57.23 55.27 83.56
25 50.63 68.07 52.34 68.06 85.34
30 48.82 69.06 47.61 62.01 87.02
35 42.09 71.64 45.14 71.21 93.45
40 51.08 67.31 59.60 78.25 94.01
45 54.05 78.27 66.07 81.37 96.08
50 53.02 86.32 77.82 84.19 97.09

The conventional approach has shown an up-and-down trend; the quickest rate is about half of the fastest rate of the 3D picture simulation. A 3D picture simulation evaluation approach is used to assess the realisticness of the terrain. The accuracy of this article’s three-dimensional picture simulation judgment rate continuously increases as the range expands. In contrast, the accuracy of the old method is always maintained at roughly 40%. Thus, it can be concluded that the 3D image simulation evaluation approach is more accurate than the old way of determining the reasonableness of landscape design. Data have been collected and analyzed in the field, as shown in the results and discussion section, which contrasts the results of the conventional method with those of this investigation. Judgment rates for the traditional methodology increase steadily from 0 to 10 but remain lower than those for the 3D simulation approach.

5 Conclusion

Various ideas and technologies are involved in planning and constructing smart gardens and landscapes in the framework of human-centered AI-based web 4.0. Humans share resources with other species and exist in this ecosystem; it’s critical to our well-being. Environmental science teaches us how to protect our natural resources and ecosystems in the face of human population development and human sources. Research in Landscape is unique that it includes novel findings and thoughtful assessments of current landscape methods. The journal attracts many academic and professional readers and a worldwide audience. Appreciate ecological and physical science principles and methodologies and how they might solve environmental problems. Appreciate environmental challenges and the connections between human and environmental systems from an ethical, cross-cultural, and historical perspective. To meet the environmental policy’s continuous improvement and pollution control aims, a company sets quantifiable environmental targets. Real scene model analysis and matching rate have been enhanced in simulations allowing for the integration of 3D images and human-centered AI-based web 4.0 technologies.

  1. Funding information: This research is supported by scientific research fund project of Guizhou University of Finance and Economics (2021X).

  2. Conflict of interest: The author declares that there’s no conflict of interest in this article.

References

[1] Liu Y, Chun OU, Yao X, Yuan H. Landscape design of hill ecology and rural human settlement environment based on the analysis of geographic information system. Arab J Geosci. 2021;14(16):1–18.10.1007/s12517-021-09248-9Search in Google Scholar

[2] Pei L. Green urban garden landscape design and user experience based on virtual reality technology and embedded network. Environ Technol Innov. 2021;24:101738.10.1016/j.eti.2021.101738Search in Google Scholar

[3] Esirgapovich JA. City parks and some issues of landscape and environmental aspect. Int J Discoveries Innov Appl Sci. 2021;1(5):145–7.Search in Google Scholar

[4] Kuittinen M, Hautamäki R, Tuhkanen EM, Riikonen A, Ariluoma M. Environmental product declarations for plants and soils: How to quantify carbon uptake in landscape design and construction? Int J Life Cycle Assess. 2021;26:1–17.10.1007/s11367-021-01926-wSearch in Google Scholar

[5] Shan P, Sun W. Research on 3D urban landscape design and evaluation based on geographic information system. Environ Earth Sci. 2021;80(17):1–15.10.1007/s12665-021-09886-ySearch in Google Scholar

[6] Jahani A, Allahverdi S, Saffariha M, Alitavoli A, Ghiyasi S. Environmental modeling of landscape aesthetic value in natural urban parks using artificial neural network technique. Mode Earth Syst Environ. 2021;8:163–72.10.1007/s40808-020-01068-2Search in Google Scholar

[7] Allahyar M, Kazemi F. Effect of landscape design elements on promoting neuropsychological health of children. Urban Forestry & Urban Green. 2021;65:127333.10.1016/j.ufug.2021.127333Search in Google Scholar

[8] Kang L. Street architecture landscape design based on Wireless Internet of Things and GIS system. Microprocess Microsyst. 2021;80:103362.10.1016/j.micpro.2020.103362Search in Google Scholar

[9] Wang M. Investigation of remote sensing image and big data analytic for urban garden landscape design and environmental planning. Arab J Geosci. 2021;14(6):1–15.10.1007/s12517-020-06304-8Search in Google Scholar

[10] Gu X, Hong T. Landscape design of road environment based on computer. Journal of physics: Conference series. Vol. 1649, IOP Publishing; 2020, September. p. 012012. No. 1.10.1088/1742-6596/1649/1/012012Search in Google Scholar

[11] Watkins H, Robinson JM, Breed MF, Parker B, Weinstein P. Microbiome-inspired green infrastructure: A toolkit for multidisciplinary landscape design. Trends Biotechnol. 2020;38(12):1305–8.10.1016/j.tibtech.2020.04.009Search in Google Scholar PubMed

[12] Liu M, Nijhuis S. Mapping landscape spaces: Methods for understanding spatial-visual characteristics in landscape design. Environ Impact Assess Rev. 2020;82:106376.10.1016/j.eiar.2020.106376Search in Google Scholar

[13] Raymond CM, Reed M, Bieling C, Robinson GM, Plieninger T. Integrating different understandings of landscape stewardship into the design of agri-environmental schemes. Environ Conserv. 2016;43(4):350–8.10.1017/S037689291600031XSearch in Google Scholar

[14] Liao C, Qiu J, Chen B, Chen D, Fu B, Georgescu M, et al. Advancing landscape sustainability science: Theoretical foundation and synergies with innovations in methodology, design, and application. Landsc Ecol. 2020;35(1):1–9.10.1007/s10980-020-00967-0Search in Google Scholar

[15] Djenontin IN, Zulu LC, Etongo D. Ultimately, what is forest landscape restoration in practice? Embodiments in sub-Saharan Africa and implications for future design. Environ Manag. 2021;68(5):619–41.10.1007/s00267-020-01360-ySearch in Google Scholar PubMed

[16] Rafi ZN, Kazemi F, Tehranifar A. Public preferences toward water-wise landscape design in a summer season. Urban Forestry Urban Green. 2020;48:126563.10.1016/j.ufug.2019.126563Search in Google Scholar

[17] Bell S, Mishra HS, Elliott LR, Shellock R, Vassiljev P, Porter M, et al. Urban blue acupuncture: A protocol for evaluating a complex landscape design intervention to improve health and wellbeing in a coastal community. Sustainability. 2020;12(10):4084.10.3390/su12104084Search in Google Scholar

[18] Cattaneo T, Giorgi E, Ni M. Landscape, architecture and environmental regeneration: A research by design approach for inclusive tourism in a rural village in China. Sustainability. 2019;11(1):128.10.3390/su11010128Search in Google Scholar

[19] Motealleh P, Moyle W, Jones C, Dupre K. Creating a dementia-friendly environment through the use of outdoor natural landscape design intervention in long-term care facilities: A narrative review. Health Place. 2019;58:102148.10.1016/j.healthplace.2019.102148Search in Google Scholar PubMed

[20] Atwa SMH, Ibrahim MG, Saleh AM, Murata R. Development of sustainable landscape design guidelines for a green business park using virtual reality. Sustain Cities Soc. 2019;48:101543.10.1016/j.scs.2019.101543Search in Google Scholar

[21] Bukhtoiarov NI, Nedikova EV. Design of environmental technologies on agricultural land. AER Advances Eng Res. 2019;182:365–8.10.2991/ciggg-18.2019.69Search in Google Scholar

[22] Trang HLT, Lee JS, Han H. How do green attributes elicit pro-environmental behaviors in guests? The case of green hotels in Vietnam. J Travel & Tour Mark. 2019;36(1):14–28.10.1080/10548408.2018.1486782Search in Google Scholar

[23] Pelsmakers S. The environmental design pocketbook. Routledge; 2019.10.4324/9780429347573Search in Google Scholar

[24] Turbé A, Barba J, Pelacho M, Mugdal S, Robinson LD, Serrano-Sanz F, et al. Understanding the citizen science landscape for european environmental policy: An assessment and recommendations. Citiz Sci Theory Pract. 2019;4(1):34.10.5334/cstp.239Search in Google Scholar

[25] Fan X, Zhou B, Wang HH. Urban landscape ecological design and stereo vision based on 3D mesh simplification algorithm and artificial intelligence. Neural Process Lett. 2021;53(4):2421–37.10.1007/s11063-021-10442-9Search in Google Scholar

[26] Araújo SO, Peres RS, Barata J, Lidon F, Ramalho JC. Characterising the agriculture 4.0 landscape—Emerging trends, challenges and opportunities. Agronomy. 2021;11(4):667.10.3390/agronomy11040667Search in Google Scholar

[27] O’Neill EG, Maravelias CT. Towards integrated landscape design and biofuel supply chain optimization. Curr Opin Chem Eng. 2021;31:100666.10.1016/j.coche.2020.100666Search in Google Scholar

[28] Kang L. Street architecture landscape design based on Wireless Internet of Things and GIS system. Microprocess Microsyst. 2021;80:103362.10.1016/j.micpro.2020.103362Search in Google Scholar

[29] Xie Z. Key factors influencing landscape design in informatized urban development. Ekoloji. 2019;28(107):3535–40.Search in Google Scholar

[30] Zhang T. Research on environmental landscape design based on virtual reality technology and deep learning. Microprocess Microsyst. 2021;81:103796.10.1016/j.micpro.2020.103796Search in Google Scholar

[31] Badach J, Raszeja E. Developing a framework for the implementation of landscape and greenspace Indicators in sustainable urban planning. Waterfront landscape management: Case studies in Gdańsk, Poznań and Bristol. Sustainability. 2019;11(8):2291.10.3390/su11082291Search in Google Scholar

[32] Liao C, Qiu J, Chen B, Chen D, Fu B, Georgescu M, et al. Advancing landscape sustainability science: Theoretical foundation and synergies with innovations in methodology, design, and application. Landsc Ecol. 2020;35(1):1–9.10.1007/s10980-020-00967-0Search in Google Scholar

[33] Fan X, Zhou B, Wang HH. Urban landscape ecological design and stereo vision based on 3D mesh simplification algorithm and artificial intelligence. Neural Process Lett. 2021;53:1–17.10.1007/s11063-021-10442-9Search in Google Scholar

[34] Li R, Xu D. Distribution of landscape architecture based on 3D images and virtual reality rationality study. IEEE Access. 2020;8:140161–70.10.1109/ACCESS.2020.3010097Search in Google Scholar

[35] Brown M, Murtha T, Wang Y, Wang L. ILAS: Intrinsic landscape assessment system for landscape design and planning in the national capital region. Journal of Digital Landscape. Architecture. 2019;2019(4):84–94.Search in Google Scholar

[36] Picchi P, van Lierop M, Geneletti D, Stremke S. Advancing the relationship between renewable energy and ecosystem services for landscape planning and design: A literature review. Ecosyst Serv. 2019;35:241–59.10.1016/j.ecoser.2018.12.010Search in Google Scholar

[37] Alpak EM, Özkan DG, Düzenli T. Systems approach in landscape design: A studio work. Int J Technol Des Educ. 2018;28(2):593–611.10.1007/s10798-017-9402-7Search in Google Scholar

[38] Seçkin NP. Environmental control in architecture by landscape design. A/Z ITU J Fac Archit. 2018;15:197–211.10.5505/itujfa.2018.90022Search in Google Scholar

Received: 2022-02-16
Revised: 2022-09-29
Accepted: 2022-11-09
Published Online: 2023-01-31

© 2023 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Downloaded on 5.3.2024 from https://www.degruyter.com/document/doi/10.1515/jisys-2022-0092/html?lang=en
Scroll to top button