Mobile QR Code QR CODE

2025

Reject Ratio

81.5%


  1. (School of Fine Arts, Huaiyin Teachers College, Huaian 223300, China)



Tourism oriented rural areas, Rural landscape, Landscape pattern, Interactive contrastive learning, GIS algorithm

1. Introduction

The vigorous development of rural leisure and comprehensive tourism, and emphasized the important value of the tourism industry in multiple speeches, determining its primary position in the five major happiness industries [1]. With the surge in demand for rural tourism, a large number of traditional agriculture and rural areas are moving towards the development path of tourism transformation, which has also promoted economic growth in rural areas, expanded employment opportunities for villagers, and become an important way to achieve rural revitalization [2, 3]. However, some regions excessively pursue the economic benefits of the tourism industry, completely disregarding the local natural background conditions, blindly developing tourism, and even directly imitating successful cases from other regions. Not only does it cause homogenization of rural scenery and tourism products, but it also disrupts the normal operation of rural ecosystems, leading to a series of ecological problems such as hindered ecological processes, decline in ecosystem service capabilities, and destruction of species diversity [4, 5]. In 2017, the “National Tourism Work Report” clearly pointed out the misconceptions of “blind development, excessive development, and predatory development” in tourism development, and the phrase “wholeheartedly fishing and damaging the environment” exposed the real situation of some rural tourism development [6].

The root cause of this phenomenon lies in a lack of understanding of the landscape science, it is not only necessary to explore the specific methods of tourism oriented rural landscape planning and design, spatial optimization, but also to grasp the laws and influencing factors of landscape evolution from a temporal perspective, attach importance to the scientific evaluation of rural landscape resources, and implement rural landscape planning, design, and construction according to local conditions. Rural planning design units in various provinces and cities have carried out a large number of village planning and design practices, and have made certain progress. However, due to a lack of profound understanding and practical experience in rural areas, the formulation of village planning is also in the exploratory stage. From the current results of village planning, there are still some problems [7, 8]. The rural planning process often relies heavily on government intentions and the experience of design units, lacking scientific evaluation of natural and cultural resources. This results in subjectivity and a focus on living and production spaces, such as construction land development and permanent farmland protection, while neglecting the ecological environment and optimization of natural spatial structures [9, 10]. A limited understanding of rural spatial dynamics leads to static planning approaches that overlook the laws of spatial development [11]. Additionally, insufficient theoretical research contributes to the lack of systematic methods in rural planning. Current academic studies often focus on landscape style, regional culture, and industry-scenery integration, which are important for rural protection and development but fail to address ecological and environmental issues [12]. Research on rural ecological landscapes remains at the theoretical level, with limited practical guidance for landscape planning. The interaction between landscape pattern and process, central to landscape ecology, is crucial for understanding landscape mechanisms, evaluating resources, and optimizing space, offering essential theoretical support for landscape planning and ecological space management [13, 14].

However, there are still shortcomings in the research and application of this theory in rural areas. It is necessary to conduct rural landscape research based on landscape ecology, grasp the spatial evolution process and laws of rural areas, and explore spatial planning and design approaches. The requirement to establish a village planning database based on GIS platform for planning results indicates that China’s rural planning and construction work is undergoing a transformation from traditional and extensive to precise and scientific planning models [15, 16]. The digitization of rural planning is not only reflected in the digitization of planning results, but also in the inevitable trend of rural spatial planning based on data analysis and evaluation. GIS technology provides technical support for the scientific evaluation, analysis, and spatial optimization of rural landscape resources [17, 18]. Firstly, identify the “ecological source area” of the Mashan Peninsula, and establish an “ecological resistance surface” in combination with the actual situation.

2. Theory and Model of Hierarchical Iterative Multimodal Interactive Learning Method

2.1. Teaching Interaction Model for Augmented Reality Integrated Learning Environment

There are two widely accepted definitions of augmented reality. They believe that augmented reality is a state between “reality” and “virtuality”. As shown in Eqs. (1) and (2), he distributes the real environment and virtual environment at both ends, while augmented reality and augmented virtual are located in the middle and included in the mixed reality scene.

(1)
$G = \sum_{k=0}^{|\tau-1|} r(s_t, a_t, s_{t+k+1}),$
(2)
$G = \sum_{k=0}^{|\tau-1|} \gamma^k r(s^t, a^t, s^{t+k+1}).$

He advocates that augmented reality technology should have three specific features: the combination of virtual and real, three-dimensional registration, and real-time interaction. Since the invention of a headshot display, people’s exploration of augmented reality technology has not stopped. As shown in Eqs. (3) and (4).

(3)
$\upsilon_\pi(s_t) = \sum_{a \in A} \pi(a_t | s_t) \times \sum_{s_{t+1} \in S} p(s_{t+1} | s_t, a_t)[r_{t+1} + \gamma \upsilon_\pi(s_{t+1})],$
(4)
$q_\pi(s_t, a_t) = \sum_{s_{t+1} \in S} p(s_{t+1} | s_t, a_t)[r_{t+1} + \gamma v_\pi(s_{t+1})].$

As shown in Eqs. (5) and (6), This provides educators with a new teaching perspective, allowing learners to interact in a more natural way, significantly enhancing their engagement. Especially in the teaching of abstract content, teaching effects that are difficult to achieve solely through image display can be achieved through augmented reality technology.

(5)
$\upsilon_\pi^T(s_t) = \sum_{a_t} \pi(a_t | s_t) \times \sum_{s_{t+1}} p(s_{t+1} | s_t, a_t)[r_{t+1} + \gamma v_\pi^{T-1}(s_{t+1})],$
(6)
$J(\pi_\theta) = E[R(\tau)]_{\pi \sim \tau}.$

Fixed equipment, can be used in museums or public places. As shown in Eqs. (7) and (8), they can be composed of a screen, a binocular with a base, or a projector, which can directly display images on physical objects. Mobile devices can be freely carried by users, and the most. Head mounted displays are becoming increasingly popular.

(7)
$\theta^* = \theta + \alpha \nabla J(\pi_\theta),$
(8)
$\nabla_\theta J(\pi_\theta) = \nabla_\theta E_{\tau \sim \pi_\theta}[R(\tau)].$

In AR devices, there are usually two types of perspective methods used, namely optical perspective based on optical principles and video perspective based on camera principles. Optical perspective equipment consists of a semitransparent screen, as shown in Eqs. (9) and (10), where information is covered on the screen and appears to be a part of the real world. Fixed transparent screens typically use optical perspective. Video perspective devices use one or more cameras to capture scenes, add virtual information to recorded images, and display it to users. Most handheld devices use video perspective.

(9)
$\nabla_\theta J(\pi_\theta) = E_{\tau \sim \pi_\theta} \left[ \sum_{t=0}^T \nabla_\theta \log \pi_\theta(a_t | s_t) R(\tau) \right],$
(10)
$\hat{g} = \frac{1}{|D|} \sum_{\tau \in D} \sum_{t=0}^T \nabla_\theta \log \pi_\theta(a_t | s_t) R(\tau).$

2.2. Hierarchical Iterative Multimodal Teaching Interaction Strategy

Three basic features of augmented reality technology were proposed, namely the combination of virtual and real, 3D registration, and real-time interaction. Augmented reality technology can achieve a combination of virtual and real effects through optical or video perspective, allowing users to clearly see the real world within their field of view while also seeing the virtual objects displayed by the device. As shown in Eqs. (11) and (12), unlike the complete immersion in virtual reality, augmented reality brings a feeling that is not detached from the real world.

(11)
$A^\pi(s, a) = Q^\pi(s, a) - V^\pi(s),$
(12)
$\theta_{new} \leftarrow \text{argmax}_\theta \tilde{\mathcal{L}}(\theta | \theta_{old}).$

The three-dimensional registration technology possessed by augmented reality technology calculates the position information through the transmitter and perceptron of sensor signals, thereby determining the position and adjusting the position of the displayed virtual object in the real environment. Real time interaction technology stands out as the interaction between devices and users, as shown in Eqs. (13) and (14). In addition to traditional interaction through input devices, it can also perform interaction in areas such as voice and action. Its efficient computation can also ensure real-time interaction.

(13)
$\text{maximize}_{\theta s, a \sim \theta_{old}} [A_{\theta_{old}}(s, a) - \beta D_{KL}(\theta_{old}, \theta)],$
(14)
$G(v, w) = \sigma(\alpha \cdot \text{heading}(v, w) + \beta \cdot \text{dist}(v, w) + \gamma \cdot \text{velocity}(v, w)).$

Unlike virtual reality presents the virtual environment to users, as shown in Eqs. (15) and (16), augmented reality technology can use virtual data augmentation to overlay virtual objects on top of the real world on the basis of displaying the real world.

(15)
$r^t = r^t_{goal} + r^t_{step} + r^t_{collision} + r^t_{guide} + r^t_{beep},$
(16)
$r^t_{new} = r^t_{old} + r^t_v + r^t_\omega.$

As shown in Eqs. (17) and (18), augmented reality technology has many advantages that other technologies cannot match, so it has broad application space and bright development prospects in the field of education. Visualizing both real and virtual content simultaneously is crucial in the field of augmented reality.

(17)
$r^t_v = \begin{cases} \omega_1 * \Delta v, & \text{if } \Delta v > 0.2, \\ \omega_2 * (v_t)^2, & \text{if } \Delta v \le 0.2, \end{cases}$
(18)
$r^t_\omega = \omega_3 * v^t_{robot} * \omega^t_{robot}.$

Use a pair of glasses to display enhanced content overlaid on the real surrounding environment. The difference between HoloLens released by Microsoft and most other similar devices is that it is a complete AR system running the Windows 10 operating system and includes a central processing unit, customized holographic processing units, various types of sensors, as shown in Eqs. (19) and (20), and optical lenses viewed using a holographic projector. Wearers can see virtual objects placed on top of the real environment while seeing the real environment, and operate through eye contact, voice, and gestures, giving people a brand new visual experience.

(19)
$D = \frac{R}{\sqrt{U^2 + V^2 + 1}},$
(20)
$L_c = \left[ \frac{\sum_{i=1}^n \Delta LU_{i-j}}{2 \sum_{i=1}^n LU_i} \right] \times \frac{1}{T} \times 100\%.$

3. Response Of Ecosystem Service Value to Changes in Landscape Pattern of Tourism Oriented Rural Areas Based on GIS Algorithms

3.1. Response Analysis Method for The Value of Ecosystem Services In Tourism Oriented Rural Areas

The HoloLens device features a color camera that adjusts resolutions to capture screenshots and videos aligned with the user’s line of sight, but it is not used for tracking or positioning [19, 20]. It also includes four grayscale tracking cameras—-two in the front with high field overlap and two on the sides with minimal overlap [21, 22]. Additionally, the device has a TOF depth-sensing camera for measuring pixel range using “long pulse” (0.8-3.5m) and “short pulse” (0-0.8m) modes [23, 24]. The short pulse is primarily for hand gesture recognition, while the long pulse is used for environmental mapping. The TOF camera’s view overlaps with the color camera but only partially [25, 26]. The HoloLens SDK retrieves camera positions using two reference systems: TOrigin, which describes the device’s position with translation components only, and TDevice, which includes both translation and orientation relative to TOrigin. Fig. 1 illustrates the interactive GIS decision tree optimization algorithm, showing how the system manages positional data in real-time applications [27, 28].

Fig. 1. Interactive GIS decision tree optimization algorithm diagram.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig1.png

To obtain absolute position through each camera, the principle of generating indoor 3D data involves using HoloLens' spatial mapping function. By switching HoloLens to developer mode and utilizing Microsoft’s project resource package, users can perform spatial scanning and reconstruct indoor 3D spaces. During this process, users must wear HoloLens, scanning in all directions indoors and briefly pausing to ensure complete data capture. The resulting spatial map allows users to see a triangular mesh that closely fits real indoor objects. Immersion theory, also known as flow theory, describes a state of deep focus on enjoyable activities, similar to how artists or athletes immerse themselves in their work. In this state, individuals find the activity itself rewarding, regardless of achieving further goals. Initially applied to online activities, immersion theory describes this state as a coherent sequence of continuous machine interactions, characterized by pleasure, selflessness, and self-motivation.

The theory comprises two core concepts: skills, representing the user’s abilities, and challenges, representing the user’s control over the environment. An imbalance between skills and challenges can lead to disinterest or frustration. To achieve immersion, a balance between these elements is essential. The immersion theory model has evolved over time, with the four-channel segmentation model advancing from a three-channel model by adding an imperceptible state. The eight-channel model further refines this by introducing additional states: relaxed (between boredom and disinterest), worried (between disinterest and anxiety), aroused (between anxiety and immersion), and controlled (between immersion and boredom). Fig. 2 illustrates the architecture of a multi-scale landscape evaluation model based on immersion theory. To experience immersion, it is crucial to achieve concentrated attention, interest, and enjoyment during experiential activities. Augmented reality (AR) technology, such as HoloLens, enhances this by helping students quickly engage in learning, stimulating their interest, and providing high-quality enjoyment through real-time interaction, thereby fostering an immersive experience. In educational contexts, “teaching interaction” is proposed to describe meaningful interactive phenomena in distance education, distinguishing it from general “interaction.” This concept allows for a more precise definition within specific teaching scenarios, emphasizing its targeted educational significance and avoiding conceptual generalization.

Fig. 2. Architecture diagram of multi-scale landscape evaluation model.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig2.png

3.2. Methods and Optimization Strategies for Constructing a Safe Landscape Pattern in Tourism Oriented Rural Areas

In this model, the teaching interaction of distance learning is divided into two levels: conceptual interaction and information interaction. In this model, the three types of teaching interactions are closely related and interact with each other, jointly promoting the construction of learner concepts. As a low-level interaction, operational interaction ensures the progress of high-level interaction activities. With the advancement of science and technology and the enrichment and improvement of media functions, proficient use of media lays the foundation for accurately and efficiently finding learning resources; Information exchange takes operational interaction as a prerequisite, and interacts with teachers, students, and teaching resources through media links in remote learning. During the process of information exchange, students, based on their own experience, integrate the results of the interaction between new and old concepts through feedback from information exchange, and ultimately regulate their own reactions. During teaching interaction, conceptual interaction belongs to top-level interaction, which occurs simultaneously during information exchange, promoting the generation and formation of new concepts. The teaching interaction model of augmented reality integrated learning environment is also divided into two levels: conceptual interaction and information interaction, both of which occur simultaneously during the class process. Fig. 3 shows the GIS assisted path optimization algorithm diagram. The underlying information interaction consists of the interaction between teachers and students, as well as the interaction between students and learning resources. The operational interaction occurs in the interaction between students and the media, namely HoloLens. These three interactions are also divided into two levels, namely the online level and the offline level.

Fig. 3. GIS assisted planning path optimization algorithm diagram.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig3.png

The offline level occurs in the operational interaction between students and the augmented reality device HoloLens. Unlike traditional online interaction that can occur at any time, the teaching interaction model based on augmented reality integrated learning environment focuses more on HoloLens augmented reality devices as a medium for students to interact with online resources, promoting conceptual interaction between teachers and students in offline face-to-face teaching, and helping students utilize online teaching resources. The hierarchical structure of this model mainly reflects two levels: online and offline. The online level represents the information interaction that students achieve with teaching resources through operational interaction with the media. This online interaction is not separated from classroom operational interaction and information interaction. It occurs during the teacher’s teaching process, utilizing the augmented reality feature of HoloLens to display virtual objects on the real world. Learners can interact with the rich online teaching resources at any time during the listening process. The offline level represents the information exchange between students and teachers, which can be observed through inquiry-based language communication and unobservable conceptual interaction. During the listening process, when the interaction between HoloLens and online resources is not sufficient to answer any questions, students can ask the teacher multiple times. The iterative nature of this model is mainly reflected in the multiple rounds of information interaction between students and teachers, as well as in the process of concept interaction. Table 1 shows the threat factors and related attributes of the Mashan Peninsula. From the model, we can know that the learning process of students is the formation process of new concepts. So, every information exchange between students and teachers is not a repetitive process of the same concept, but a process of gradually forming new concepts based on specific concepts of students under the influence of teaching resources. Each concept exchange is an iterative spiral upward.

Table 1. Threat factors and related attributes of the Mashan Peninsula.

Threat factors Maximum impact distance/km Weight
Land used for building 0.2 0.8
Cultivated land 0.03 0.6
Naked ground 0.02 0.2
Other land use 0.08 0.4

The multimodality of this model refers to multimodal interaction, mainly reflected in the operational interaction between students and the media. As an augmented reality device, HoloLens not only connects rich online teaching resources, but also provides multi-sensory interaction functions that cannot be provided by mobile phones, tablets, and other devices. Through various sensory channels such as vision, hearing, and action, human-computer interaction can be carried out. The multimodal interaction process can provide students with a tangible experience and achieve a more natural interaction effect. The application of GIS technology in rural spatial analysis abroad is relatively comprehensive, including spatial pattern changes, spatial planning, ecological environment protection, biodiversity protection, agricultural production, cultural heritage protection, and various social issues. However, the application of domestic GIS technology in spatial analysis of rural areas is not comprehensive enough. The main research content focuses on the distribution characteristics, evolution process, influencing factors of settlements, and spatial optimization based on suitability evaluation. There are relatively many studies on this type of topic, and clear research ideas and methods have been formed, which have been widely applied.

Combined with natural geographic data such as altitude, slope, water source, and roads, the distribution of rural settlements in typical low mountain. Based on Voronoi plot coefficient of variation and landscape pattern index, the distribution characteristics of rural settlements in the semi-arid agricultural pastoral transitional zone of Inner Mongolia were analyzed, and the main influencing factors of rural settlement distribution were explored by combining factors such as water system, terrain, and topography. Fig. 4 shows the evaluation map of rural landscape types, analysis functions of remote sensing data using GIS technology. And discuss its influencing factors and driving mechanisms from the perspectives of nature, humanities, and society; Use the nearest neighbor statistics, Ripley’s K function, Based on kernel density analysis, explore the spatial distribution characteristics and influencing factors of rural settlements in Weixian from the Shang and Zhou dynasties to the Qing dynasty. By utilizing GIS technology for spatial analysis and data processing, combined with the multi factor comprehensive evaluation method, the suitability evaluation of rural land use can be carried out, which is conducive to promoting the rationality and scientific of rural spatial development and utilization.

Fig. 4. Rural landscape type evaluation map.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig4.png

4. Research on Rural Landscape Design Based on Comparative Learning Interactive GIS Algorithm

The landscape pattern index is a key method in landscape ecology for quantitative analysis, reflecting characteristics from various angles through simple calculations. Landscape pattern changes are crucial to understanding the interaction between landscape processes and the ecological environment. Although many studies have examined large-scale landscapes and their relationship to ecological issues like natural disasters and biodiversity, such research offers limited guidance for spatial planning in small and medium-sized rural areas. In particular, tourism-oriented rural regions lack in-depth analysis of landscape patterns. This study focuses on the Mashan Peninsula, a tourism-oriented rural area in Wuxi City. By comparing the landscape patterns before and after the development of tourism, the study explores the changes in landscape structure over time. The coupling of tourism development and rural landscape changes provides insights into the impact of tourism activities, offering valuable guidance for rural landscape planning, ecological protection, and sustainable development. Table 2 highlights the carbon density parameters of various landscape types, illustrating the environmental impacts.

Table 2. Carbon density parameters of different landscape types.

Landscape type Aboveground carbon Underground carbon Soil carbon Total
Land used for building 1.20 4.35 4.35 9.90
Broad leaved forest land 9.03 24.69 27.42 61.14
Shuitiandi 4.77 17.19 16.04 38.00
Shrubs and grasslands 5.67 14.38 16.78 36.83
Naked ground 0.01 4.29 7.46 11.76
Water system 0.06 7.90 8.11 16.07
Other land use 1.20 4.35 4.35 9.90

The dynamic degree of land use assesses the intensity of landscape changes over time, using two indicators: “single land use dynamic degree” and “comprehensive land use dynamic degree.” This study employs the “single land use dynamic degree” to analyze changes in specific landscape types on the Mashan Peninsula. A lower absolute value indicates stability, while higher values reflect intense changes. The landscape pattern index, typically calculated using Fragstats, quantifies landscape characteristics at the landscape and type levels, excluding patch-level data due to its low relevance for regional analysis. This study focuses on key indices like landscape area, morphology, and distribution. The Mashan Peninsula comprises seven landscape types: construction land, broad-leaved forest, cultivated land, shrubs and grasslands, bare land, water bodies, and others. Broad-leaved forests dominate the island with the largest area and best connectivity but are decreasing and fragmenting over time. Meanwhile, construction land expands significantly, transitioning from a complete state to fragmentation and back toward completion.

From 1984 to 2020, construction land and shrubs in the Mashan Peninsula increased, while broad-leaved forests, water bodies, bare land, and other land types decreased. Cultivated land remained stable. Construction land saw the largest increase (+192.1%), particularly from 1991 to 2010. Shrubs and grasslands grew by 62.5%, with the most significant growth from 1991 to 2001. Broad-leaved forests experienced the sharpest decline, followed by a partial recovery, while water bodies also shrank. Bare land, despite its small base, saw a sharp reduction (-84.9%). By calculating land use dynamics across four periods (1984-2020), the study reveals that from 1984 to 1991, the overall landscape change was minimal, with a comprehensive dynamic degree of 1.64%, indicating relative stability. Bare land had the highest dynamic increase (+30.92), while cultivated land (-3.17%) and shrubs (-2.93%) showed significant reductions. Fig. 5 illustrates soil quality and vegetation cover evaluation. The dynamics of other landscape types remained stable during this period.

Fig. 5. Soil quality and vegetation coverage assessment map.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig5.png

From 1991 to 2001, the comprehensive land use dynamic degree was relatively similar to the previous period, at 1.63%. The three types of landscapes, namely shrubbery, construction land, and cultivated land, exhibit positive dynamics. The attitude towards shrubbery is the highest (10.94%), followed by construction land (5.54%) and cultivated land (5.04%), indicating that the growth dynamics of shrubbery scale are more intense at this time. The broad-leaved forest and bare land showed negative dynamics, with results of -2.6% and -2.7%, respectively, indicating a trend of reduction in the area of these two types of landscapes. The lowest dynamic calculation result of water landscape indicates that its state is relatively stable. The comprehensive land use dynamics showed a rapid increase, increasing to 1.99%, indicating that the regional landscape dynamics were more intense and the landscape pattern had undergone significant changes. Among them, the three types of landscapes, namely construction land, cultivated land, and shrubs and grasslands, show positive dynamics. The dynamic value of construction land is the highest (4.6%), and the increase is the largest. Cultivated land (1.85%) and shrubs and grasslands (1.57%) are second to construction land, indicating that the expansion trend of construction land in the Mashan Peninsula is the most obvious at this time, and cultivated land and shrubs and grasslands are also increasing. The dynamic value of bare ground is negative (-2.55%), indicating its decreasing trend.

From 2010 to 2020, the comprehensive land use dynamics in the Mashan Peninsula increased slightly to 2.04%. Broad-leaved forests and construction land both showed positive dynamics, with broad-leaved forests growing at a slightly higher rate (2.5%) than construction land (2.0%). Farmland, shrub, and grassland landscapes exhibited negative dynamics, with the largest decrease in bare land (-9.2%), followed by farmland (-2.7%) and shrubs/grassland (-2.4%). Water landscapes remained stable with the lowest dynamic value. The landscape type transfer matrix highlights key transitions, such as water bodies transforming into cultivated land and bare land transitioning into construction land. Broad-leaved forests, farmland, and shrubs also showed notable conversions into bare land.

The development of ecosystem service research has evolved from global assessments of ecosystem functions–such as supply, regulation, support, and cultural services–into a focus on the interaction between human activities and ecosystem services. However, studies have primarily focused on macro scales, neglecting rural areas and the role of landscape structure in ecosystem service efficiency. There is a need for research on ecosystem services at a rural scale to better inform landscape planning and improve ecosystem management practices in these regions.

5. Experimental Analysis

In economics, sensitivity coefficient analysis is the degree to which changes in certain key factors lead to changes in the net present value of investment plans, while ensuring other conditions remain unchanged. Fig. 7 shows the evaluation of transportation network density and accessibility, reflecting the sensitivity of net present value to changes in key factors. Similarly, in the study of ecosystem services, the unit area ecosystem service value of various landscapes represents the key factor, while the total value of ecosystem services represents the net present value of investment plans.

Fig. 6. Water distribution and water quality assessment chart.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig6.png

Fig. 8 shows the population density and evaluation of public service facilities. The sensitivity coefficient can verify the accuracy of the “ecosystem service value per unit area” of the landscape. If the sensitivity coefficient is less than 1, it indicates that the ecosystem service value has no elasticity, and the calculation results have high credibility. Anti regularization indicates that the ecosystem service value has elasticity, but the accuracy of the calculation results is insufficient.

This study adjusted the “unit area ecosystem service value” of different types of landscapes in the Mashan Peninsula by 50%, evaluating the sensitivity of the total regional. Fig. 9 shows the evaluation of economic activities and employment opportunities. To further reveal the interaction mechanism between landscape pattern changes and ecosystem services in the Mashan Peninsula, a quantitative model.

Fig. 7. Traffic network density and accessibility evaluation chart.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig7.png

Fig. 8. Population density and evaluation of public service facilities.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig8.png

Fig. 9. Economic activity and employment opportunity assessment chart.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig9.png

In the regression model, the absolute value of the standardized regression coefficients for each landscape index represents the explanatory power of the index for ecosystem service value. The higher the value, the greater the impact of the landscape index on ecosystem service value, indicating that the corresponding landscape pattern characteristics have a strong impact on ecosystem services. Therefore, based on this method, identify the main landscape pattern characteristics that affect the ecosystem service value of the Mashan Peninsula, and use this as a basis to grasp the landscape pattern optimization approach aimed at enhancing the efficiency of ecosystem services. The support service function is most significantly influenced by the overall diversity of the landscape in the Mashan Peninsula, the complexity of shrub and grassland shapes, and the scale of broad-leaved forests. Fig. 10 shows the evaluation of cultural heritage sites and tourism resources.

Secondly, as shrubs and grasslands expand and spread towards broad-leaved forests, the continuity of the broad-leaved forest landscape is also disrupted, the overall connectivity of the landscape decreases, and the ecosystem service capacity declines. The cultural service function is most significantly affected by the overall diversity of the landscape in the Mashan Peninsula, as well as the morphology of construction land and shrubland patches. Fig. 11 shows the energy consumption and carbon emissions assessment chart. The complexity of construction land and shrub and grassland forms within the region indicates that it is spreading towards broad-leaved forests, causing a serious decrease in the aesthetic value of broad-leaved forests.

Fig. 10. Cultural heritage sites and tourism resource evaluation map.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig10.png

Fig. 11. Energy consumption and carbon emission assessment chart.

../../Resources/ieie/IEIESPC.2026.15.4.554/fig11.png

6. Conclusions

The Mashan Peninsula’s landscape pattern has continuously changed with the development of tourist resorts, showing an upward trend in dynamics. Construction land has expanded, encroaching on surrounding landscapes, particularly broad-leaved forests, leading to a fragmented, heterogeneous landscape with reduced connectivity and dominance. Factors driving these changes include tourism construction, agricultural demand, and policy promotion. As tourism develops, the value of ecosystem services in the region–covering supply, regulation, support, and cultural services–has declined, with regulation services most affected. There is a significant correlation between the ecosystem service value and landscape pattern indices, though different indices impact ecosystem services differently. Correlation and regression analysis help identify key grid characteristics affecting ecosystem services, allowing for pattern optimization strategies to improve efficiency. The study includes five typical rural areas, each covering 50 square kilometers, with seven primary land use types analyzed.

Detailed GIS data collection and analysis yielded 350 data points across five regions. Agricultural land averages 40% of the total area, followed by forest land at 30%, construction land at 15%, water areas at 10%, and other land types at 5%. The average vegetation coverage is 65%, ranging from a high of 78% to a low of 52%. The average slope is 15 degrees, with flat terrain (less than 5 degrees) accounting for 20%, medium slope (5-20 degrees) for 50%, and steep terrain (greater than 20 degrees) for 30%. Water bodies average 12% distribution, with the largest covering 1.5 square kilometers and the smallest just 0.2 square kilometers. To promote tourism-oriented rural landscapes, it’s essential to balance ecological functions with the protection of local cultural heritage. Constructing distinct “ecological security” and “cultural landscape security” patterns can ensure ecological stability and safeguard cultural resources. By overlaying these patterns for comprehensive zoning and strategic layout, we can support the coordinated development of ecological protection and cultural initiatives, offering valuable insights for sustainable landscape planning in tourism-focused rural areas.

References

1 
L. Chen , Z. Zhang , J. Wu , X. Wei , W. Bai , Dynamic retrieval model of quantitative data of power grid resources based on 3D geographic information systems (GIS), Journal of Nanoelectronics and Optoelectronics, Vol. 17, No. 2, pp. 344-350, 2022DOI
2 
S. Chen , Q. Guo , L. Li , Sustainable land use dynamic planning based on GIS and symmetric algorithm, Advances in Civil Engineering, Vol. 2022, Art. no. 4087230, 2022DOI
3 
H. Cho , Memory-efficient flow accumulation using a look-around approach and its OpenMP parallelization, Environmental Modelling & Software, Vol. 167, Art. no. 105771, 2023DOI
4 
S. Dasgupta , A. Baral , A. Lahiri , Optimization of electrode-spacer arrangement using simplex algorithm, IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 30, No. 2, pp. 726-733, 2023DOI
5 
B. Decardi-Nelson , J. Liu , An efficient implementation of graph-based invariant set algorithm for constrained nonlinear dynamical systems, Computers & Chemical Engineering, Vol. 164, Art. no. 107906, 2022DOI
6 
C. Ding , Z. Wang , J. Li , Y. Ding , Structural optimization design of three phase enclosed GIS disconnector based on MOGOA and ELM, IEEJ Transactions on Electrical and Electronic Engineering, Vol. 19, No. 2, pp. 215-221, 2024DOI
7 
W. Fang , G. Chen , W. Li , M. Xu , W. Xie , C. Chen , W. Wang , Y. Zhu , A PRPD-based UHF filtering and noise reduction algorithm for GIS partial discharge, Sensors, Vol. 23, No. 15, Art. no. 6763, 2023DOI
8 
P. G. P. S. Fernandes , A. E. Arantes , E. F. Nobre Júnior , Optimization model for earthwork allocations considering the construction of multiple haul roads: GIS-based integrated approach, Journal of Construction Engineering and Management, Vol. 150, No. 4, Art. no. 05024001, 2024DOI
9 
R. He , J. Yan , D. Zhao , L. Lu , Y. Geng , Gas-insulated switchgear partial discharge acoustic-electric joint localisation method based on the Salp swarm algorithm and least squares estimation, Measurement, Vol. 225, Art. no. 114020, 2024DOI
10 
Z. Ijaz , C. Zhao , N. Ijaz , Z. U. Rehman , A. Ijaz , M. F. Junaid , Geospatial modeling of heterogeneous geotechnical data using conventional and enhanced conception of modified Shepard method-based IDW algorithms: Application and appraisal, Bulletin of Engineering Geology and the Environment, Vol. 82, No. 11, Art. no. 428, 2023DOI
11 
J. Ji , H. Cui , T. Zhang , J. Song , Y. Gao , A GIS-based tool for probabilistic physical modelling and prediction of landslides: GIS-FORM landslide susceptibility analysis in seismic areas, Landslides, Vol. 19, No. 9, pp. 2213-2231, 2022DOI
12 
X. Jiang , Q. Wang , Y. Long , S. Zhang , Y. Fang , D. Hu , Vibration signal features prediction of GIS equipment based on improved slime mold optimization algorithm optimizing CNN-BiLSTM, Traitement du Signal, Vol. 39, No. 4, pp. 1109-1117, 2022DOI
13 
J. Jiao , H. Yang , Z. Chen , Y. Gu , J. Liu , Inversion of mine transient electromagnetic data via the PSO-GIS algorithm from a conical source, IEEE Transactions on Geoscience and Remote Sensing, Vol. 60, pp. 1-12, 2022DOI
14 
S. Karimi , I. Iordanova , Integration of BIM and GIS for construction automation: A systematic literature review (SLR) combining bibliometric and qualitative analysis, Archives of Computational Methods in Engineering, Vol. 28, No. 7, pp. 4573-4594, 2021DOI
15 
B. Kotyra , High-performance watershed delineation algorithm for GPU using CUDA and OpenMP, Environmental Modelling & Software, Vol. 160, Art. no. 105613, 2023DOI
16 
B. Kotyra , Ł. Chabudziński , Fast parallel algorithms for finding the longest flow paths in flow direction grids, Environmental Modelling & Software, Vol. 167, Art. no. 105728, 2023DOI
17 
P. Lemenkova , Artificial intelligence for computational remote sensing: Quantifying patterns of land cover types around Cheetham Wetlands, Port Phillip Bay, Australia, Journal of Marine Science and Engineering, Vol. 12, No. 8, Art. no. 1279, 2024DOI
18 
J. Li , X. Pei , X. Wang , D. Yao , Y. Zhang , Y. Yue , Transportation mode identification with GPS trajectory data and GIS information, Tsinghua Science and Technology, Vol. 26, No. 4, pp. 403-416, 2021DOI
19 
T. Li , X. Wang , S. Zeng , C. Gu , C. Dong , L. Zhen , D. Zhang , H. Liu , Gas insulated switchgear partial discharge type identification based on frequency division features and improved long short-term memory neural networks, Journal of Nanoelectronics and Optoelectronics, Vol. 19, No. 2, pp. 169-177, 2024DOI
20 
Z. Li , Y. Qian , H. Wang , X. Zhou , G. Sheng , X. Jiang , Partial discharge fault diagnosis based on Zernike moment and improved bacterial foraging optimization algorithm, Electric Power Systems Research, Vol. 207, Art. no. 107854, 2022DOI
21 
F. Lin , H.-P. Hsieh , Multicriteria route planning for in-operation mass transit under urban data, Applied Sciences, Vol. 12, No. 6, Art. no. 3127, 2022DOI
22 
C. Liu , C. Hu , C. Zhao , Y. Sun , T. Xie , H. Wang , Research on urban storm flood simulation by coupling K-means machine learning algorithm and GIS spatial analysis technology into SWMM model, Water Resources Management, Vol. 38, No. 6, pp. 2059-2078, 2024DOI
23 
L. Liu , The use of GIS-based genetic algorithm in water pollution control planning, Desalination and Water Treatment, Vol. 241, pp. 223-229, 2021DOI
24 
M. Liu , GIS-based urban agglomeration landscape dynamic observation and simulation prediction algorithm, Journal of Sensors, Vol. 2022, Art. no. 8117539, 2022DOI
25 
X. Liu , Y. Li , L. T. Sizemore , X. Xie , J. Wu , A deep-learning approach to detect and classify heavy-duty trucks in satellite images, IEEE Transactions on Intelligent Transportation Systems, Vol. 25, No. 10, pp. 13323-13338, 2024DOI
26 
F. Lv , K. Ma , H. Geng , S. Wang , Z. Li , Q. Zhang , Influence of surface charge on distribution characteristics of metal particle dust in DC GIS/GIL, IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 30, No. 5, pp. 2412-2420, 2023DOI
27 
M. Shen , GIS positioning based on optical sensing detection and simulation of tourism route planning in rural tourism scenic areas, Optical and Quantum Electronics, Vol. 56, No. 3, Art. no. 284, 2024DOI
28 
W. Sun , H. Ma , S. Wang , Application of SCNGO-VMD-SVM in identification of gas insulated switchgear partial discharge, IEEE Access, Vol. 12, pp. 43838-43848, 2024DOI
29 
W. Sun , H. Ma , S. Wang , A novel fault diagnosis of GIS partial discharge based on improved whale optimization algorithm, IEEE Access, Vol. 12, pp. 3315-3327, 2024DOI
30 
Y. Sun , Y. Hua , L. Mou , X. X. Zhu , CG-Net: Conditional GIS-aware network for individual building segmentation in VHR SAR images, IEEE Transactions on Geoscience and Remote Sensing, Vol. 60, pp. 1-15, 2022DOI
Chuanhui Qiu
../../Resources/ieie/IEIESPC.2026.15.4.554/au1.png

Chuanhui Qiu received his bachelor’s degree in literature from Shenyang Aerospace University in 2004, a master’s degree in engineering from Jiangnan University in 2013. He is currently working as a lecturer in the Academy of Arts,Huaiyin Normal University. His research areas and directions include Protection and utilization of rural landscape environment and so forth.