Spatial and Temporal Evolution of Urban Green Space Pattern Based on GIS Sensors and Remote Sensing Information: Taking Xi’an as an Example

Abstract

In order to solve the practical application of the continuously developing remote sensing technology in urban planning, this paper proposes a method of temporal and spatial evolution of urban green spatial pattern based on GIS remote sensing information. Based on the Landsat Image data of the main urban area of Xi’an from 2000 to 2012, different classification methods are used to extract the urban green space information and compare the accuracy. The classification results with high accuracy are selected to analyze the temporal and spatial evolution law of urban green space and the change of landscape pattern in the study area. In this paper, the change of vegetation coverage can be divided into five levels: significant degradation: <-0.006; slight degradation: -0.006~-0.002; stable: -0.002~0.002; slight improvement: 0.002~0.006; and significant improvement: >0.006. The results of this paper prove that this method can be used to understand and evaluate the ecological consequences of urbanization and improve our quality of life. At the same time, it can provide basic information for decision-making.

This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Introduction

As a part of urban ecosystem, urban green space plays an important role in urban sustainable development because many ecosystem services and functions are crucial to urban ecological integrity and human well-being [1]. Urban green space is an important part of urban ecosystem. It has important ecological auxiliary function for urban development and progress. It is an important indicator of urban ecological civilization construction and plays an important role in regulating urban ecological environment [2].

With the acceleration of urbanization and urban expansion, the area of urban green space continues to shrink and gradually becomes a scarce resource. However, people’s attention to green living standards continues to increase with the growth of economic development level [3]. Urban green space has become the focus of many disciplines [4].

On the one hand, urban green space provides healthy and comfortable rest places for urban residents. On the other hand, it is also an important measure for the government to reasonably layout the urban spatial structure, and promoting its healthy development has become an important content of urban planning [5]. Urban green space improves the quality of life and air quality, reduces the use of refrigeration and heating energy, purifies the urban environment, regulates the temperature and humidity of the air, and makes the urban environment more aesthetically desirable [6]. With the continuous development of urban economy and the acceleration of urbanization, the environmental problems faced by cities are becoming more and more serious, such as urban heat island effect, air pollution, water resources pollution, ecological environment damage, excessive population expansion, and serious damage to biodiversity. They have seriously affected the physical and mental health of urban residents and are one of the most severe challenges faced by human society in the 21st century. Rational planning of urban green space can effectively solve the severe challenges faced in the process of urban development in China [7, 8]. At the same time, green space plays an important role in enterprise location, local employment rate, and tourism development, as shown in Figure 1.

[figure(s) omitted; refer to PDF]

2. Literature Review

Urban green space is the key to the coupling of urban spatial development and ecological environment, and it is also the key to effectively alleviate the problem of environmental degradation in urbanization areas. As an important living urban infrastructure form attached to urban land resources, urban green space is an important part of urban ecosystem with white and clean function, and it is also an important symbol to measure the level of urban sustainable development and civilization [9]. Due to the differences in the stages of urban development and the focus of ecological and environmental protection research at home and abroad, scholars at home and abroad have different definitions and understandings of green space. Foreign countries mostly refer to urban open space [10]. In 1877, the concept of urban open space was first put forward in London, England [11]. In 1906, urban open space was formally defined as “any enclosed or unclosed land, in which less than 1 / 20 of the land without buildings or buildings, and the rest of the land was used as parks or entertainment, or as waste areas, or unused areas” [12].

Urban open green space emphasizes the openness and greenness of space. All countries in the world are committed to building various forms of urban open green space, such as urban parks, corridors, and green belts. As a kind of urban land, urban green space is distributed in the urban area from the perspective of comprehensive consideration of urban land utilization and has a relatively clear land scope. It pays more attention to the landscape aesthetic benefits, social benefits, and ecological environment benefits of green space. Urban green space is treated from the perspective of city and large region. Its composition and classification have the characteristics of comprehensiveness and integrity. It includes not only the types of green space divided by urban green space but also the agricultural production land, water body, mountain, grassland, wetland, and so on in the whole region. It emphasizes the coordinated development of urban structure and the sustainability and rationality of the development of urban spatial structure, including all land within the city that has a direct or indirect impact on the improvement of urban ecological environment and people’s life [13]. According to the classification standard of urban green space issued by the Ministry of Housing and Urban-Rural Development in 2017, urban green space is divided into five categories: park green space, square green space, protective green space, auxiliary green space, and regional green space [14]. The traditional concept of green space in China actually refers to green space in a narrow sense.

Zhu et al. define urban green space as a green network system composed of garden green space, urban forest, three-dimensional space greening, urban farmland, and water wetland [15]. Liu et al. pointed out that urban green space contains at least the following two meanings: first, urban regional space based on natural acquisition, mainly covered by artificial green vegetation; the second is to provide various social service functions, including ecology, entertainment, and aesthetics [16]. Zhu et al. believe that urban green space refers to the area covered by green plants inside and around the city, including natural or man-made green space such as grassland, forest land, and cultivated land [17]. Wang et al. put forward that urban green space refers to the space covered by living plants in the city, which is the sum of urban forests, crops, shrubs, and other plants. It includes natural vegetation (although the natural vegetation of most cities has disappeared), seminatural vegetation, and artificial vegetation, mainly including the central area of the city and its surrounding areas [18]. Zhang and Shen believe that urban green space can include any vegetation in the urban environment, including parks, open spaces, residential gardens, or street trees. These places provide residents with contact space, opportunities for leisure, and entertainment, as well as habitats of natural species, and maintain biodiversity [19].

Therefore, based on the satellite data with long time series and medium spatial resolution, this paper applies remote sensing and GIS technology to extract the green spatial information of the main urban area of Xi’an and discusses the extraction, composition structure, and temporal and spatial evolution characteristics of the green spatial information in the study area, so as to provide reference for the sustainable development of the city.

3. Research Methods

3.1. Elements of Green Space in Xi’an

Urban green space classification is the basis of green space information extraction, space-time evolution of green space, and landscape pattern analysis [20]. In this paper, according to the classification of land use status, the research of scholars and the actual situation of the main urban area of Xi’an, and according to the contribution of the constituent elements of urban green space to green space, the spatial types of the study area are divided into green space and nongreen space, in which green space includes urban green space cultivated land, and water body, and nongreen space includes construction land and other land, as shown in Table 1.

Table 1

Classification of urban green space.

Urban space type	Land use type	Types and characteristics of green space
Green space	Cultivated land	Land for planting crops, including paddy field, dry land, and irrigated land
	Urban green space	Forest land mainly refers to the land where trees, bamboos, and shrubs growGrassland includes natural grassland, swamp grassland, artificial grassland, and other grasslands
	Water body	Including rivers (canals), lakes, reservoirs, pits and ponds, coastal and inland beaches
Nongreen space	Land used for building	Including residential land, commercial land, industrial and mining storage land, public management and public service land, special land, transportation land, hydraulic construction land, etc.
	Other land	Including bare land, sandy land, bare rock gravel land, saline alkali land, etc.

3.2. Xi’an Green Space Information Extraction Method

Image classification is an integral part of image processing. It extracts topic information by learning the relationship between spectral features and various categories or topics of interest to users. It is a complex process and needs to consider a variety of factors [21]. As an important part of urban ecosystem, urban green space has the function of ecological balance and is closely related to human life. Its integrity, diversity, and systematicness are disturbed by human activities to a great extent, showing dynamic characteristics. Quantitative analysis of the type composition, quantitative characteristics, dynamic evolution, and other characteristics of green space can help understand the development and evolution process of green space and provide crucial basic data and analysis basis for the analysis of the ecological and environmental effects caused by urbanization and human activities [22]. User demand, the scale of the research area, economic conditions, and the skills of analysts are important factors affecting the selection of remote sensing data, the design of classification program, and the quality of classification results [23].

3.2.1. Maximum Likelihood Method

Maximum likelihood classification (MLC) is a statistical method of pattern recognition [24]. For each pixel in the image, the probability of belonging to a certain category is calculated and assigned to the category with the highest probability. Generally, assuming that the distribution of each class in the multidimensional space is Gaussian distribution, the mean and covariance matrix of maximum likelihood classification are obtained from the training samples and used to effectively model the class [25]. The basis of this assumption is Bayesian decision rule (BDR), in which different kinds of probabilities need to be specified. Using BDR, we must have a priori knowledge of different kinds of probabilities. Maximum likelihood classification is a common method in supervised classification. It is suitable for most image processing software with high classification accuracy. It is the most commonly used parameter classifier in practice, as shown in $$\begin{array}{}\text{(1)}& g_{i}x=\ln p{\omega }_i-\frac{1}{2\ln {\sum }_i}-\frac{1}{2{x-m_i}^T{\Sigma }_i^{-1}x-m_i},\end{array}$$where $i$ is class $i$; $x$ is n-dimensional data, where $n$ is the number of bands; $p{\omega }_i$ is the probability that a certain class $i$ appears in the image, and it is assumed that all classes are the same; ${\sum }_i$ is the determinant of the covariance matrix of some type $i$ data; ${\Sigma }_i^{-1}$ is the inverse of a class $i$ covariance matrix; $m_i$ is the average vector of some $i$.

3.2.2. Support Vector Machine Classification

Support vector machine (SVM) was developed in the mid-1990s. This method is a supervised classification method based on statistical learning theory and structural risk minimization (SRM) principle. SVM not only achieves good results in pattern recognition, function estimation, and regression analysis but also shows many unique advantages in solving small sample and nonlinear and high-dimensional pattern recognition. The simplest way to train support vector machines is to use linear classifiers. The formula is expressed as $$\begin{array}{}\text{(2)}& X_i,y_i,i=1,2,\cdots ,k.\end{array}$$

In formula (2), $k$ is the number of training samples: $X$ is $N$-dimensional space, $X\in RN$; $y$ is a class label, and $y\in -1,+1$. If there is a vector $W$ perpendicular to the linear hyperplane (determining the direction of the discrimination plane) and a scalar $b$ showing the offset of the discrimination hyperplane from the origin, these classes are considered to be linearly separable. Two hyperplanes can be used to distinguish data points in various classes. The formula is expressed as $$\begin{array}{}\text{(3)}& \begin{array}{c}W{X}_i+b\ge +1,y=+1,\hfill \\ W{X}_i+b\le -1,y=-1.\hfill \end{array}\end{array}$$

3.2.3. Random Forest Classification

The basic idea of random forest classification method is as follows: firstly, bootstrap sampling technology is used to extract training samples from the original data set, and each training sample accounts for about two-thirds of the original data set. Then, a classification regression tree is established for each training sample to generate a forest composed of $N$$\text{CART}$ decision trees. In the growth process of each tree, $m$ characteristic variables are randomly selected from all $M$ characteristic variables as the input sample of each node, $m\le M$. Finally, the classification results of each decision tree are synthesized by voting, and the classification result with the largest number of votes is the final result. The out-of-bag tree is called one-third of the out-of-bag tree. The out-of-bag samples are used as the test data to generate the decision tree. This part of the out-of-bag data is used to estimate the internal error, resulting in OOB error. The random forest classification method uses Gini index as the attribute selection measure. For a given training set $T$, randomly select a pixel and indicate that it belongs to a class $Ci$, $f{C}_i,T/T$ is the probability that the selected pixel belongs to class $Ci$. The Gini index can be expressed as $$\begin{array}{}\text{(4)}& \text{Gini}\text{Index}=\sum \sum _{j\ne i}\frac{f{C}_i,T}{T}\frac{f{C}_{i}T}{T}.\end{array}$$

3.3. Research Ideas

Based on the five Landsat Image data of 1994, 2000, 2006, 2013, and 2018, taking Xi’an land use map, high-resolution Google map, and administrative division vector data as auxiliary data, this paper carries out a series of preprocessing for Landsat Image data, such as radiometric calibration, atmospheric correction, geometric correction, and image clipping. For the preprocessed data, the maximum likelihood classification method, support vector machine classification method, and random forest classification are used to extract the green space information of the study area, and the accuracy is verified to obtain the optimal green space extraction method: the random forest classification method and landscape index method are selected for landscape pattern analysis. The research idea of this paper is shown in Figure 2:

[figure(s) omitted; refer to PDF]

4. Result Analysis

Using MODIS NDVI data with stronger timeliness compared with common remote sensing images, this paper analyzes the vegetation cover change and distribution pattern in Xi’an from 2000 to 2012, so as to provide decision-making reference for the evaluation and adjustment of Xi’an industrial pattern and economic layout.

The MODIS NDVI/EVI MOD13Q1 data provided by NASA used in this paper has a spatial resolution of 250 m and a temporal resolution of 16d. Affected by the climatic environment and vegetation growth in Xi’an, and according to the existing research results, July to September of each year is defined as the growth season of plants in this area. This paper selects the relevant data of plant growth season images in Xi’an from 2000 to 2012 for correlation analysis.

In order to eliminate the influence of cloud as far as possible, 117 MODIS NDVI images from 2000 to 2012 were synthesized by maximum value composite in years. In ArcGIS10.0, select the max command to maximize the processing, and then obtain the data related to NDVI-Max in 13 growing seasons from 2000 to 2012.

4.1. Study on Vegetation Cover Change

The interannual linear change trend of NDVI is used to judge the nature and intensity of vegetation cover change. On the basis of each pixel, NDVI-Max is linearly fitted, and the trend slope is calculated by the least square method, as shown in the following: $$\begin{array}{}\text{(5)}& b=\frac{x_i-y}{x_i-x}.\end{array}$$In the equation, $b$ is the trend of the slope, $x$ and $y$ are the year and the NDVI-Max value of the year respectively, and $x$ and $y$ are the mean value of the year and the NDVI-Max value from 2000 to 2012, respectively. A positive slope indicates the increase of vegetation coverage. The larger the value, the more obvious the trend of vegetation coverage increase. A negative slope indicates a decline in vegetation coverage. The more obvious the value, the more obvious the trend of vegetation coverage reduction.

In order to determine the natural fluctuation range of NDVI-Max in the study area from 2000 to 2012, combine the basic geography.

According to the data, the water body without vegetation cover is extracted in the whole area of Xi’an. Considering that the resolution of remote sensing image is 250 m, 10 patches with an area greater than 1 km² are selected as the experimental area. By analyzing the linear trend slope of the average value of NDVI-Max in the growing season from 2000 to 2012, it is found that almost all tribes are in the range of -0.002~0.002. As shown in Figure 3, set this range as the natural fluctuation amplitude of vegetation cover in Xi’an, and extend the limit of natural fluctuation outward twice to obtain 0.006. Take this value as the limit of slight change and significant change of vegetation cover. Therefore, the change of vegetation coverage can be divided into five levels: significant degradation: <-0.006; slight degradation: -0.006~-0.002; stable: -0.002~0.002; slight improvement: 0.002~0.006; and significant improvement: >0.006.

[figure(s) omitted; refer to PDF]

4.2. Interannual Variation of Vegetation Coverage in Xi’an

Use ArcGIS10.0 to draw the distribution map of vegetation cover change trend in Xi’an, and count the area of each vegetation cover change category, and calculate the proportion, as shown in Table 2. The results show that the vegetation cover in Xi’an has been mainly stable and increased in recent 13 years, accounting for 81.42%. The vegetation coverage in some areas is seriously reduced and concentrated, mainly distributed in the northeast of Yanbian County, the river area of Miyi County, the west, the area under the jurisdiction of Xi’an city, and the south of Xi’an city. As the distribution is relatively concentrated, it may be caused by the continuous development of industrial land and construction land.

Table 2

Statistics of covered mounting categories of each value.

Vegetation cover change category	The measure of area	Proportion
Significant reduction	420	6.23%
Stable	1020	14%
Increase	2300	34%

In order to compare the changes of NDVI-Max in different growth seasons, the annual growth season NDVI-Max of two increase types and two decrease types are counted, respectively, and their change trend charts are drawn, as shown in Figures 4(a) and 4(b).

[figure(s) omitted; refer to PDF]

Figures 4(a) and 4(b) show that NDVI-Max increased by an average of 0.01 per year in the growth season of the significantly improved type in the past 13 years (the confidence is 99.9%), the annual average increase rate is 152%, and the type of slight improvement is NDVI-Max in the growing season.

The average annual increase is 0.0038 (confidence is 98.0%), and the average annual increase rate is 0.49%. The average annual reduction rate of NDVI-Max in the growth season of significant degradation type is 0.0103 (confidence 99.9%), and the annual average reduction rate is 1.35%. The average annual reduction rate of NDVI-Max in the growth season of slight degradation type is 0.0036 (confidence 99.9%), and the annual average reduction rate is 0.45%.

Make statistics on the changes of the average value of NDVI-Max in the growing season from 2000 to 2012 in the whole Xi’an area, and draw a broken line diagram of the changes. As shown in Figure 5, it can be seen that the overall vegetation coverage in Xi’an is good, and the NDVI changes above 0.76. According to the change trend of NDVI, the vegetation cover change in the past 13 years can be divided into two periods: the fluctuation and rise period from 2000 to 2009 and fluctuation decline period from 2009 to 2012. The reasons for this change also need to be studied in combination with various data.

[figure(s) omitted; refer to PDF]

5. Conclusion

This paper uses ENVI5.3, ArcGIS10.0, and other related software, based on the high-resolution remote sensing image data of Xi’an, taking Dadukou area as an example, through the splicing, cutting, and extracting green space information of remote sensing images. In a series of steps, the production of thematic map of urban green space landscape types was completed, and the database of urban green space system was established. On this basis, the spatial structure of green space landscape in Dadukou area is analyzed by using statistical analysis software such as Patch Analyst, GS+9.0, and SPSS, mainly from the aspects of urban green space landscape patch composition, urban green space landscape patch scale grade, urban green space landscape patch fragmentation, urban green space landscape overall index, urban green space landscape spatial heterogeneity, and so on. At the same time, based on the MODIS NDVI data of Xi’an, this paper studies the vegetation cover change and distribution pattern in this area from 2000 to 2012. The conclusions are as follows:

(1) The landscape types of green space in Dadukou area of Xi’an are mainly divided into 9 types: urban park, community park, protective green space, residential green space, public construction green space, road green space, scenic and recreational green space, river green space, and other ecological green space. Through the calculation of the number of various green space types, the number of various green space landscape patches, the area of various green space landscape, the maximum patch index, and other indexes, the composition analysis of green space landscape patches is realized. The results show that the total area of green space in Dadukou area of Xi’an is 219.35 hectares, with a total of 2565 patches, and the overall green space coverage rate of this area reaches 38.3%. Among all types of green space, public construction green space and residential green space account for the absolute advantage in the total area of green space, and the corresponding number of patches is also the largest. The area of road green space is the least, but the number of patches is large, which is mainly due to the fact that there are many intersections on the road and the green space is mostly distributed on both sides of the road, so it is difficult to form continuous green patches. The number of protective green space patches is the least, but the area is large, mainly because the protective green space is mostly distributed in flakes and the patch area is large. According to the analysis results of the maximum patch index of different green space landscape types, it can be concluded that the maximum patch index of other green spaces is the maximum, indicating that the green space type is uneven in area distribution

(2) Through the analysis of the scale level of green landscape patches, it is found that the green patches in Dadukou area of Xi’an are mainly small patches, which are scattered. Small patches ($\text{area}<0.1$ hm²) are mainly residential green space and public construction green space; medium patches (0.1 hm²~0.5 hm²) are mainly residential green space and public construction green space; larger patches (0.5 hm²~1 hm²) are mainly public construction green space; large patches ($\text{area}>1$ hm²) are mainly protective green space

(3) The results show that the fragmentation degree of road green space is the most serious, while the fragmentation degree of protective green space is the lowest. At the same time, it also verified the relationship between the evaluation patch area, patch density, and the degree of green space fragmentation; that is, the larger the average patch area, the smaller the patch density, and the lower the degree of green space fragmentation

References

[1] J. Jin, S. R. Sheppard, B. Jia, C. Wang, "Planning to practice: impacts of large-scale and rapid urban afforestation on greenspace patterns in the Beijing plain area," Forests, vol. 12 no. 3,DOI: 10.3390/f12030316, 2021.

[2] H. Dang, J. Li, Y. Zhang, Z. Zhou, "Evaluation of the equity and regional management of some urban green space ecosystem services: a case study of main urban area of Xi'an city," Forests, vol. 12 no. 7,DOI: 10.3390/f12070813, 2021.

[3] S. Heo, A. Nori-Sarma, S. Kim, J. T. Lee, M. L. Bell, "Do persons with low socioeconomic status have less access to greenspace? Application of accessibility index to urban parks in Seoul, South Korea," Environmental Research Letters, vol. 16 no. 8,DOI: 10.1088/1748-9326/ac12f1, 2021.

[4] F. Salamone, B. Barozzi, L. Danza, M. Ghellere, I. Meroni, "Correlation between indoor environmental data and biometric parameters for the impact assessment of a living wall in a ZEB lab," Sensors, vol. 20 no. 9,DOI: 10.3390/s20092523, 2020.

[5] S. Liu, X. Zhang, Y. Feng, H. Xie, Z. Lei, "Spatiotemporal dynamics of urban green space influenced by rapid urbanization and land use policies in Shanghai," Forests, vol. 12 no. 4,DOI: 10.3390/f12040476, 2021.

[6] P. Cheng, M. Min, W. Hu, A. Zhang, "A framework for fairness evaluation and improvement of urban green space: a case of Wuhan metropolitan area in China," Forests, vol. 12 no. 7,DOI: 10.3390/f12070890, 2021.

[7] F. Li, R. Wang, S. Lu, M. Shao, J. Ding, Q. Sun, "Spatiotemporal simulation of green space by considering socioeconomic impacts based on A SD-CA model," Forests, vol. 12 no. 2,DOI: 10.3390/f12020202, 2021.

[8] S. Dinda, N. D. Chatterjee, S. Ghosh, "An integrated simulation approach to the assessment of urban growth pattern and loss in urban green space in Kolkata, India: a GIS-based analysis," Ecological Indicators, vol. 121 no. 2021, 2021.

[9] Y. Zhang, T. Zhang, Y. Zeng, C. Yu, S. Zheng, "The rising and heterogeneous demand for urban green space by Chinese urban residents: evidence from Beijing," Journal of Cleaner Production, vol. 313 no. 3,DOI: 10.1016/j.jclepro.2021.127781, 2021.

[10] Y. Sun, X. Wang, J. Zhu, L. Chen, J. Wu, "Using machine learning to examine street green space types at a high spatial resolution: application in Los Angeles county on socioeconomic disparities in exposure," Science of the Total Environment, vol. 787 no. 7,DOI: 10.1016/j.scitotenv.2021.147653, 2021.

[11] P. Kowe, O. Mutanga, T. Dube, "Advancements in the remote sensing of landscape pattern of urban green spaces and vegetation fragmentation," International Journal of Remote Sensing, vol. 42 no. 10, pp. 3797-3832, DOI: 10.1080/01431161.2021.1881185, 2021.

[12] Y. Yao, S. Zhang, Y. Shi, M. Xu, J. Zhao, "Landscape pattern change of impervious surfaces and its driving forces in shanghai during 1965–2010," Water, vol. 13 no. 14,DOI: 10.3390/w13141956, 2021.

[13] C. C. Nicholson, N. M. Williams, "Cropland heterogeneity drives frequency and intensity of pesticide use," Environmental Research Letters, vol. 16 no. 7,DOI: 10.1088/1748-9326/ac0a12, 2021.

[14] R. Wang, J. H. Kim, M. H. Li, "Predicting stream water quality under different urban development pattern scenarios with an interpretable machine learning approach," Science of the Total Environment, vol. 761 no. 4, 2020.

[15] Z. Zhu, B. Liu, H. Wang, M. Hu, "Analysis of the spatiotemporal changes in watershed landscape pattern and its influencing factors in rapidly urbanizing areas using satellite data," Remote Sensing, vol. 13 no. 6,DOI: 10.3390/rs13061168, 2021.

[16] C. Liu, F. Zhang, V. C. Johnson, P. Duan, H. T. Kung, "Spatio-temporal variation of oasis landscape pattern in arid area: human or natural driving?," Ecological Indicators, vol. 125 no. 1,DOI: 10.1016/j.ecolind.2021.107495, 2021.

[17] C. Zhu, X. Zhang, M. Zhou, S. He, K. Wang, "Impacts of urbanization and landscape pattern on habitat quality using OLS and GWR models in Hangzhou, China," Ecological Indicators, vol. 117 no. 3,DOI: 10.1016/j.ecolind.2020.106654, 2020.

[18] S. Wang, Y. Zhao, Z. Xu, G. Yang, N. Ma, "Landscape pattern and economic factors' effect on prediction accuracy of cellular automata-Markov chain model on county scale," Open Geosciences, vol. 12 no. 1, pp. 626-636, DOI: 10.1515/geo-2020-0162, 2020.

[19] J. Zhang, H. Shen, "Geo-spatial analysis and optimization strategy of park green space landscape pattern of garden city– a case study of the central district of Mianyang city Sichuan province," European Journal of Remote Sensing, vol. 53 no. 1, pp. 309-315, DOI: 10.1080/22797254.2020.1725788, 2020.

[20] Y. Liu, X. Cao, T. Li, "Influence of accessibility on land use and landscape pattern based on mapping knowledge domains: review and implications," Journal of Advanced Transportation, vol. 2020 no. 3,DOI: 10.1155/2020/7985719, 2020.

[21] X. Ren, C. Li, X. Ma, F. Chen, H. Wang, A. Sharma, G. S. Gaba, M. Masud, "Design of multi-information fusion based intelligent electrical fire detection system for green buildings," Sustainability, vol. 13 no. 6,DOI: 10.3390/su13063405, 2021.

[22] J. Jayakumar, S. Chacko, P. Ajay, "Conceptual implementation of artificial intelligent based E-mobility controller in smart city environment," Wireless Communications and Mobile Computing, vol. 2021,DOI: 10.1155/2021/5325116, 2021.

[23] L. Xin, M. Chengyu, Y. Chongyang, "Power station flue gas desulfurization system based on automatic online monitoring platform," Journal of Digital Information Management, vol. 13 no. 6, pp. 480-488, 2015.

[24] R. Huang, P. Yan, X. Yang, "Knowledge map visualization of technology hotspots and development trends in China’s textile manufacturing industry," IET Collaborative Intelligent Manufacturing, vol. 3 no. 3, pp. 243-251, DOI: 10.1049/cim2.12024, 2021.

[25] Q. Zhang, "Relay vibration protection simulation experimental platform based on signal reconstruction of MATLAB software," Nonlinear Engineering, vol. 10 no. 1, pp. 461-468, DOI: 10.1515/nleng-2021-0037, 2021.