Snow avalanche potential hazard area identification method and system based on multi-factor terrain constraint

Abstract

This invention provides a method and system for identifying potential avalanche hazard areas based on multi-factor topographic constraints, belonging to the field of avalanche prevention technology. Based on a digital elevation model, this invention integrates multiple topographic factors such as slope, curvature, and roughness. Through multi-factor joint screening, candidate areas with basic conditions for avalanche initiation are obtained. Combined with vegetation cover data, vegetation zones with significant anti-slip effects are eliminated. Furthermore, using connected component analysis, ridge structure identification, slope aspect consistency judgment, and internal elevation difference analysis, the candidate areas are geometrically segmented and structurally decomposed to obtain small-scale slope units that better conform to avalanche initiation patterns. Finally, scale constraints are set according to avalanche development characteristics to generate a map of potential avalanche hazard areas. This invention can extract potential avalanche hazard areas in large-scale mountainous areas, significantly improving the accuracy of avalanche hazard identification and providing a reliable data foundation for avalanche dynamics simulation, risk assessment, and protective engineering design.

Description

Technical Field

[0001] This invention belongs to the field of avalanche prevention technology, specifically relating to a method and system for identifying potential avalanche hazard areas based on multi-factor terrain constraints. Background Technology

[0002] Avalanches are one of the most significant geological hazards in high-altitude, cold regions, and are widely distributed in areas with high altitudes, steep slopes, and significant snow and ice accumulation. Avalanches are characterized by their suddenness, high speed, and great destructive power, posing a serious threat to transportation routes, power transmission facilities, settlements, and engineering projects.

[0003] Traditional techniques for identifying potential avalanche hazard zones typically employ several topographical factors, such as slope, aspect, and elevation, to assess avalanche susceptibility. These methods usually calculate macroscopic avalanche-prone areas using empirical weights or statistical models, often covering entire mountainsides or large slopes. However, avalanches exhibit significant localization and initiation unit characteristics; their initiation areas are often limited to localized slope sections on a scale of tens to hundreds of meters, rather than entire high-risk areas. Existing methods struggle to pinpoint the specific locations that might trigger avalanches within large-scale hazard zones, hindering the deployment of avalanche disaster prevention and engineering control measures.

[0004] Furthermore, traditional avalanche zoning methods have limited utilization of topographic details, typically relying only on slope or simple topographic factor thresholds, failing to fully leverage information such as slope morphology, curvature, roughness, and local geomorphic structure contained in high-resolution digital elevation models (DEMs). In complex glacier retreat zones, valley-type terrain, and slopes with alternating undulations, existing methods easily misidentify ridges, steep walls, rock ridges, and wind-eroded surfaces as potential avalanche initiation zones, resulting in numerous false alarms. Meanwhile, vegetation zones in high-altitude areas have a significant inhibitory effect on avalanche initiation.

[0005] Therefore, it is necessary to provide a method and system for identifying potential avalanche hazard areas based on multi-factor terrain constraints to solve the above problems. Summary of the Invention

[0006] This invention provides a method and system for identifying potential avalanche hazard areas based on multi-factor terrain constraints. By combining constraints from multiple factors such as slope, curvature, roughness, and vegetation, it can effectively eliminate large-scale erroneous areas and obtain local hazard units that conform to the actual initiation characteristics of avalanches. It can quickly and automatically extract potential avalanche hazard areas in large-scale complex mountainous terrain, providing basic data support for subsequent avalanche dynamics simulation, disaster impact assessment, and engineering protection measure design, thereby effectively solving at least one of the technical problems involved in the background art.

[0007] To solve the above-mentioned technical problems, the present invention is implemented as follows:

[0008] A method for identifying potential avalanche hazard zones based on multi-factor terrain constraints includes the following steps:

[0009] Step S1: Obtain the digital elevation model of the target area, and preprocess the digital elevation model to obtain continuous and complete terrain data;

[0010] Step S2: Calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness.

[0011] Step S3: Based on the terrain factors, preliminary screening of candidate areas with the potential for avalanche initiation is conducted.

[0012] Step S4: Obtain vegetation cover data of the target area, align the vegetation cover data with the terrain data, use the vegetation cover area as a mask, and remove the masked areas from the candidate areas.

[0013] Step S5: Perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units.

[0014] Step S6: Cut each slope unit according to the ridge structure, slope aspect difference and internal elevation difference to obtain multiple sub-slope units;

[0015] Step S7: Based on the engineering scale constraints of avalanche initiation, the area of ​​the sub-slope units is determined, and sub-slope units with an area smaller than the area threshold are removed to obtain the potential avalanche hazard zone.

[0016] As a preferred improvement, in step S1, the digital elevation model is derived from remote sensing imagery, and the preprocessing is performed using ArcGIS. The preprocessing process includes:

[0017] (1) Fill in abnormal elevation values ​​and void pixels;

[0018] (2) Check the spatial reference information to ensure that the coordinate system is consistent.

[0019] As a preferred improvement, in step S2, the slope is obtained by calculating the first derivative of the elevation information of the target area; the curvature is obtained by calculating the second derivative of the elevation information of the target area; and the roughness is obtained by statistically analyzing the elevation information of the target area using neighborhood normal vectors.

[0020] As a preferred improvement, the preliminary screening process for candidate regions includes the following steps:

[0021] (1) Eliminate terrain that is too steep or too gentle by adjusting the slope;

[0022] (2) Ridges and steep convex surfaces are removed by curvature removal;

[0023] (3) Remove broken rock surfaces or discontinuous areas on the ground surface by adjusting the roughness.

[0024] As a preferred improvement, the terrain factors of the candidate region satisfy the following constraints:

[0025] The slope range is 27°-48°, the curvature threshold is 7, and the roughness threshold is 0.05.

[0026] As a preferred improvement, step S5, "connectivity analysis," specifically includes the following process:

[0027] The spatial adjacency relationship of pixels is analyzed by the eight-neighbor spatial adjacency discrimination method. All continuously distributed candidate pixels are aggregated into multiple independent connected domains, forming multiple independent slope units with complete surface continuity.

[0028] As a preferred improvement, step S6 specifically includes the following process:

[0029] (1) Cutting based on ridge structure: Identify ridge units in the region that are continuous with convex features, have a large lateral slope and are long and narrow, and cut the slope units at the ridge to remove false candidate regions that extend along the ridge.

[0030] (2) Cutting based on slope aspect difference: Statistically analyze the main slope aspect of the area. If the difference between the local slope aspect and the main slope aspect exceeds the slope aspect difference threshold, then cut the area at the location of the slope aspect change.

[0031] (3) Cutting based on internal elevation difference: Analyze elevation changes along the main slope of the region. If the elevation difference exceeds the elevation difference threshold, cut the region along the contour line at that location.

[0032] As a preferred improvement, the aspect difference threshold is 45°; the elevation difference threshold is 200m.

[0033] As a preferred improvement, in step S7, the area threshold is 400㎡.

[0034] A system for identifying potential avalanche hazard zones based on multi-factor terrain constraints, comprising:

[0035] The data acquisition module is used to acquire the digital elevation model of the target area and preprocess the digital elevation model to obtain continuous and complete terrain data.

[0036] The terrain factor calculation module is used to calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness.

[0037] The preliminary screening module is used to initially screen candidate areas with the possibility of avalanche initiation based on the terrain factors.

[0038] The vegetation removal module is used to acquire vegetation cover data of the target area, align the vegetation cover data with the terrain data, use the vegetation cover area as a mask, and remove the masked area from the candidate area.

[0039] The connected component analysis module is used to perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units.

[0040] The cutting module is used to cut each slope unit according to the ridge structure, slope aspect differences and internal elevation differences to obtain multiple sub-slope units;

[0041] The discrimination module is used to perform area discrimination on the sub-slope units based on the engineering scale constraints of avalanche initiation, and remove sub-slope units with an area smaller than the area threshold to obtain the potential avalanche hazard zone.

[0042] Compared with the prior art, the present invention has the following beneficial effects:

[0043] (1) Significantly improve identification accuracy: By combining multiple factors such as slope, curvature, roughness, and vegetation, large-area error areas can be effectively eliminated, and local hidden danger units that conform to the actual avalanche initiation characteristics can be obtained.

[0044] (2) Fully automated processing: It does not rely on human experience to set complex rules, and can achieve automated and refined batch recognition in a wide range of mountainous areas. It is suitable for high-altitude and hard-to-access areas.

[0045] (3) High adaptability and versatility: This method is based on the basic features of the terrain and does not rely on meteorological measurement data. It can run stably under various terrain types and DEMs of different resolutions. By extracting micro-scale initiation units from macro-prone areas, it can quickly and automatically extract potential avalanche hazard areas in a large range of complex mountains, providing basic data support for subsequent avalanche dynamics simulation, disaster impact assessment and engineering protection measures design. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0047] Figure 1 A flowchart of the avalanche potential hazard zone identification method based on multi-factor terrain constraints provided by the present invention;

[0048] Figure 2 The results of terrain slope calculation for Example 1;

[0049] Figure 3 The results of terrain curvature calculation are from Example 1;

[0050] Figure 4 The results of terrain roughness calculation are shown in Example 1;

[0051] Figure 5 This is the result of the avalanche potential hazard zone delineation in Example 1. Detailed Implementation

[0052] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] like Figures 1-5 As shown, this embodiment provides a method for identifying potential avalanche hazard areas based on multi-factor terrain constraints, including the following steps:

[0054] Step S1: Obtain the digital elevation model of the target area, and preprocess the digital elevation model to obtain continuous and complete terrain data.

[0055] A digital elevation model (DEM) of the target area is obtained through on-site drone aerial photography or high-precision remote sensing image interpretation, with a resolution of 1-5m, and the appropriate scale is selected according to the engineering requirements.

[0056] The digital elevation model (DEM) includes an elevation matrix and its spatial reference information.

[0057] The preprocessing process includes:

[0058] (1) Fill in abnormal elevation values ​​and void pixels;

[0059] (2) Check the spatial reference information to ensure that the coordinate system is consistent;

[0060] The preprocessing process is completed using ArcGIS, which ensures the stability and continuity of subsequent terrain analysis.

[0061] Step S2: Calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness.

[0062] The slope is obtained by calculating the first derivative of the elevation information of the target area; the curvature is obtained by calculating the second derivative of the elevation information of the target area; the roughness is obtained by statistically analyzing the elevation information of the target area using neighborhood normal vectors. The specific calculation process is as follows:

[0063] (1) Slope S, used to characterize the slope angle, is an important controlling factor for avalanche initiation and is calculated as follows:

[0064] ;

[0065] In the formula, z For elevation, x East-west horizontal coordinates y The coordinates are horizontal, running north-south. / and / These represent the elevation change rates of the DEM in the east-west and north-south directions, respectively, used to calculate the slope angle.

[0066] (2) Curvature C, including planar curvature and profile curvature, is used to identify typical landform units such as ridges, valleys, and steep walls. The calculation process is shown below:

[0067] ;

[0068] In the formula, curvature C reflects the concave and convex characteristics of the slope, with positive values ​​corresponding to convex surfaces (ridges) and negative values ​​corresponding to concave surfaces (valleys).

[0069] (3) Roughness This reflects the degree of local surface undulation and is used to exclude areas such as broken rocks and cliffs that are unfavorable for stable snow accumulation. The calculation process is shown below:

[0070] ;

[0071] In the formula, Let be the slope normal vector of the i-th pixel. VRM represents the number of neighboring pixels; a larger VRM indicates higher surface roughness.

[0072] The above factors are calculated using spatial derivatives and regular neighborhood analysis to ensure the integrity of the terrain feature representation.

[0073] Step S3: Based on the terrain factors, preliminary screening of candidate areas with the potential for avalanche initiation is conducted.

[0074] Candidate region Represented as:

[0075]

[0076] In the formula: This represents the initial candidate region, which is a set of regions where the slope S, curvature C, and roughness VRM all satisfy a certain threshold range. This represents the discriminant function (or screening rule) based on the above factor threshold constraints, used to extract a set of regions that meet the conditions from the study area.

[0077] First, based on the slope threshold, identify the slope sections that may have conditions for avalanche initiation to obtain the initial screening area with a certain steepness;

[0078] Curvature information was then introduced, and by identifying typical landform features such as ridgelines, strong ridges, and steep cliffs, areas that are not conducive to snow accumulation and initiation were removed from the initial screening results.

[0079] Further utilizing roughness to reflect the continuity of the slope surface, areas with significantly high roughness, such as fractured rock surfaces, collapse walls, and strongly disturbed slopes, are excluded, making the candidate areas more consistent with the terrain morphology requirements where avalanches may occur.

[0080] The specific threshold conditions need to be determined based on the analysis of historical avalanche cases in the study area. For example, a slope range of 27°-48°, a curvature threshold of 7, and a roughness threshold of 0.05 can be selected.

[0081] Through the above-mentioned step-by-step screening, the present invention is able to preserve areas with continuous terrain, smooth slopes, and basic conditions for avalanche initiation in large-scale mountainous areas, forming the preliminary scope of potential avalanche hazard zones.

[0082] Step S4: Using remote sensing imagery, obtain vegetation cover data for the target area, align the vegetation cover data with the terrain data, and use the vegetation cover area as a mask to eliminate the masked areas in the candidate area.

[0083] Vegetation raster data was reprojected and aligned with the DEM spatial reference system to ensure correspondence between the two at the same coordinate system and resolution. Based on the mechanical relationship between snow cover and vegetation, vegetation zones (especially dense shrub and arbor areas) typically possess significant anti-slip and interception effects, significantly hindering snowfall and reducing the likelihood of avalanches, thus lacking the conditions to form avalanche initiation surfaces. After applying vegetation mask constraints to candidate areas, the vegetation-covered portions were removed, retaining only potentially unstable surface units (e.g., bare rock, bare soil, or sparsely vegetated slopes) as potential avalanche hazard zones for subsequent connectivity analysis and geometric structure cutting.

[0084] Step S5: Perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units.

[0085] The candidate regions, after being processed by vegetation masking, are input into the connected component analysis module. Through 8-neighbor spatial adjacency discrimination, all continuously connected candidate pixels within the region are clustered and labeled, aggregating all continuously distributed candidate pixels into multiple independent connected components, forming basic slope units with complete surface continuity. The analysis results automatically divide the candidate region into several independent slope units, each maintaining topographic continuity and overall geometric consistency, providing a basic processing unit for subsequent automatic segmentation based on internal geometric features (ridge structure, aspect consistency, and internal elevation differences).

[0086] Step S6: Cut each slope unit according to the ridge structure, slope aspect difference and internal elevation difference to obtain multiple sub-slope units.

[0087] In complex mountainous environments, some slope regions obtained from connected domain analysis may exhibit a large-scale continuous distribution, often containing multiple topographic units and lacking geometric consistency to serve as a single avalanche initiation zone. To obtain slope units that better reflect actual avalanche development conditions, topographic structure segmentation is performed on each connected region, specifically including the following three types of operations:

[0088] (1) Cutting based on ridge structure: Ridge areas are usually narrow, elongated "wall-like units" that extend along contour lines. They are often located near watershed boundaries and do not have the conditions for avalanche initiation. When such ridges or ridge-shaped slopes exist in the area, they are used as topographic barriers and the connected areas are cut at their boundaries;

[0089] (2) Aspect-based segmentation: Real avalanche initiation surfaces typically exhibit relatively consistent aspect characteristics. When the aspect variation within a region is excessive, it usually indicates that the region is composed of a combination of multiple different slopes and should not be judged as a whole. This invention performs statistical analysis on the aspect within each region to determine aspect consistency.

[0090] Calculation area main slope aspect For each pixel slope Calculate the deviation If there are continuous segments that cause the deviation to exceed the preset angle range, it is determined to be a slope discontinuity area. Cutting is carried out at the slope change point or slope boundary line. In this embodiment, the slope difference threshold is 45°.

[0091] (3) Cutting based on internal elevation differences: Some areas may have excessively large elevation differences, which does not conform to the actual situation. Therefore, it is necessary to set an elevation difference threshold to cut and divide the connected areas. This invention performs a one-dimensional scan of the elevation changes along the main direction within the area. When the elevation difference exceeds the threshold, it cuts along the contour line direction. In this embodiment, the elevation difference threshold is set to 200m.

[0092] Through the above-mentioned multi-factor integrated segmentation mechanism, areas that are too large, unevenly extended, or have significant internal topographic differences can be subdivided into multiple small-scale slope units with consistent structures, making the resulting units more consistent with the actual topographic characteristics of avalanche initiation.

[0093] Step S7: Based on the engineering scale constraints of avalanche initiation, the area of ​​the sub-slope units is determined, and sub-slope units with an area smaller than the area threshold are removed to obtain the potential avalanche hazard zone.

[0094] After completing the regional terrain segmentation, engineering scale constraints are applied to the resulting sub-regions to conform to the typical spatial scale characteristics of avalanche initiation units. Based on the empirical scale of avalanche initiation areas, the areas of each sub-region are screened, removing fragmented units that are too small to constitute an actual initiation surface, and retaining only areas that meet the minimum area requirement as potential hazard zones. In this embodiment, fragmented areas with an area less than 400㎡ are removed. Finally, vector boundaries are extracted from the retained areas to form the final avalanche potential hazard zone map.

[0095] This process exhibits good adaptability and scalability under different mountainous conditions, DEM resolutions, and data source conditions.

[0096] This embodiment also provides a system for performing the above-described method for identifying potential avalanche hazard areas based on multi-factor terrain constraints, including:

[0097] The data acquisition module is used to acquire the digital elevation model of the target area and preprocess the digital elevation model to obtain continuous and complete terrain data.

[0098] The terrain factor calculation module is used to calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness.

[0099] The preliminary screening module is used to initially screen candidate areas with the possibility of avalanche initiation based on the terrain factors.

[0100] The vegetation removal module is used to acquire vegetation cover data of the target area, align the vegetation cover data with the terrain data, use the vegetation cover area as a mask, and remove the masked area from the candidate area.

[0101] The connected component analysis module is used to perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units.

[0102] The cutting module is used to cut each slope unit according to the ridge structure, slope aspect differences and internal elevation differences to obtain multiple sub-slope units;

[0103] The discrimination module is used to perform area discrimination on the sub-slope units based on the engineering scale constraints of avalanche initiation, and remove sub-slope units with an area smaller than the area threshold to obtain the potential avalanche hazard zone.

[0104] It should be understood that the embodiments in this invention are only used to illustrate the technical solutions of this invention, and are not intended to limit this invention. Those skilled in the art can make modifications or equivalent substitutions to the step sequence, factor selection method, region division strategy, formula expression structure, and specific parameter settings of the implementation methods without departing from the spirit and substance of this invention, and these modifications or substitutions should all be considered to fall within the protection scope of this invention.

[0105] Example 1

[0106] This embodiment uses the avalanche potential hazard zone identification method based on multi-factor terrain constraints provided by the present invention to identify avalanche potential hazard zones in a target area. The slope of the target area is 0-87°. The calculation results are as follows: Figure 2 As shown, the terrain curvature is -8.2 to 10, and the calculation results are as follows. Figure 3 As shown, the roughness is 0-0.6, and the calculation results are as follows. Figure 4 As shown, the final map of the potential avalanche hazard area is as follows: Figure 5 As shown.

[0107] The steps described in this invention can be reduced, adjusted, or combined according to application requirements; the various technical modules described in this invention can also be implemented through software, hardware, or a combination of both. This invention emphasizes an overall process for automatic identification of potential avalanche hazard areas based on multi-factor terrain constraints. This process has good adaptability and scalability under different mountainous conditions, different resolution DEMs, and different data sources.

[0108] Those skilled in the art will understand after reading this specification that the multi-factor terrain screening concept, slope aspect consistency analysis, regional geometric cutting mechanism, and scale constraint strategy of this invention can be flexibly adjusted according to different engineering scenarios without affecting the core innovative content of this invention. In summary, any content not explicitly defined in this specification should not be construed as limiting the scope of this invention.

Claims

1. A method for identifying potential avalanche hazard zones based on multi-factor terrain constraints, characterized in that, Includes the following steps: Step S1: Obtain the digital elevation model of the target area, and preprocess the digital elevation model to obtain continuous and complete terrain data; Step S2: Calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness. Step S3: Based on the terrain factors, preliminary screening of candidate areas with the potential for avalanche initiation is conducted. Step S4: Obtain vegetation cover data of the target area, align the vegetation cover data with the terrain data, use the vegetation cover area as a mask, and remove the masked areas from the candidate areas. Step S5: Perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units. Step S6: Cut each slope unit according to the ridge structure, slope aspect difference and internal elevation difference to obtain multiple sub-slope units; Step S7: Based on the engineering scale constraints of avalanche initiation, the area of ​​the sub-slope units is determined, and sub-slope units with an area smaller than the area threshold are removed to obtain the potential avalanche hazard area. Step S6 specifically includes the following process: (1) Cutting based on ridge structure: Identify ridge units in the region that are continuous with convex features, large transverse slope and narrow shape, and cut the slope units at the ridge to remove false candidate regions extending along the ridge. (2) Cutting based on slope aspect difference: Statistically analyze the main slope aspect of the area. If the difference between the local slope aspect and the main slope aspect exceeds the slope aspect difference threshold, then cut the area at the location of the slope aspect change. (3) Cutting based on internal elevation difference: Analyze elevation changes along the main slope of the region. If the elevation difference exceeds the elevation difference threshold, cut the region along the contour line at that location.

2. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, In step S1, the digital elevation model originates from remote sensing imagery, and the preprocessing process is completed using ArcGIS. The preprocessing process includes: (1) Fill in abnormal elevation values ​​and void pixels; (2) Check the spatial reference information to ensure that the coordinate system is consistent.

3. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, In step S2, the slope is obtained by calculating the first derivative of the elevation information of the target area; the curvature is obtained by calculating the second derivative of the elevation information of the target area; and the roughness is obtained by statistically analyzing the elevation information of the target area using neighborhood normal vectors.

4. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, The initial screening process for candidate regions includes the following steps: (1) Eliminate terrain that is too steep or too gentle by adjusting the slope; (2) Ridges and steep convex surfaces are removed by curvature removal; (3) Remove broken rock surfaces or discontinuous areas on the ground surface by adjusting the roughness.

5. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 4, characterized in that, The terrain factors of the candidate region satisfy the following constraints: The slope range is 27°-48°, the curvature threshold is 7, and the roughness threshold is 0.

05.

6. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, In step S5, the "connectivity analysis" specifically includes the following process: The spatial adjacency relationship of pixels is analyzed by the eight-neighbor spatial adjacency discrimination method. All continuously distributed candidate pixels are aggregated into multiple independent connected domains, forming multiple independent slope units with complete surface continuity.

7. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, The threshold for slope aspect difference is 45°; the threshold for elevation difference is 200m.

8. The method for identifying potential avalanche hazard zones based on multi-factor terrain constraints according to claim 1, characterized in that, In step S7, the area threshold is 400㎡.

9. A system for implementing the avalanche potential hazard zone identification method based on multi-factor terrain constraints as described in any one of claims 1-8, characterized in that, include: The data acquisition module is used to acquire the digital elevation model of the target area and preprocess the digital elevation model to obtain continuous and complete terrain data. The terrain factor calculation module is used to calculate the terrain factors of the target area based on the terrain data. The terrain factors include slope, curvature, and roughness. The preliminary screening module is used to initially screen candidate areas with the possibility of avalanche initiation based on the terrain factors. The vegetation removal module is used to acquire vegetation cover data of the target area, align the vegetation cover data with the terrain data, use the vegetation cover area as a mask, and remove the masked area from the candidate area. The connected component analysis module is used to perform connectivity analysis on the candidate regions after masking to obtain multiple independent slope units. The cutting module is used to cut each slope unit according to the ridge structure, slope aspect differences and internal elevation differences to obtain multiple sub-slope units; The discrimination module is used to perform area discrimination on the sub-slope units based on the engineering scale constraints of avalanche initiation, and remove sub-slope units with an area smaller than the area threshold to obtain the potential avalanche hazard zone.

Images