Illegal fishing optimization method based on base map features

Abstract

The invention discloses an illegal fishing optimization method based on base map features, and belongs to the technical field of fishing identification, and the method specifically comprises the steps: obtaining the geographic information data of a target water area, and generating a base map feature space-time continuous field containing a water area boundary dynamic topological structure, a hydrological chromatography parameter distribution diagram and a historical illegal activity mark thermodynamic diagram; establishing a channel curvature adaptive manifold mapping coordinate system, and converting ship trajectory latitude and longitude coordinates into a curvature arc length parameter sequence under the manifold mapping coordinate system; calculating a normal penetration distance between a ship track point and a protection area unit, and detecting a space overlapping area between a track line and a sediment boundary; the staying period distribution characteristics of the ship in the high-density area are extracted, and the threshold value and the weight are corrected in combination with the real-time meteorological and hydrological data flow; an environment self-adaptive risk field gradient map is formed, a multi-stage law enforcement response instruction is output, and a monitoring equipment space deployment scheme is optimized; according to the invention, accurate identification, dynamic monitoring and efficient disposal of illegal fishing behaviors are realized.

Description

Technical Field

[0001] This invention relates to the field of fishing identification technology, and specifically to an optimization method for illegal fishing based on base map features. Background Technology

[0002] In recent years, with increased global efforts to protect fishery resources and heightened awareness of aquatic ecological environment protection, precise regulation of illegal fishing has become a crucial issue in the management of marine and inland waters. Illegal fishing not only disrupts the sustainable development balance of fishery resources but can also negatively impact the integrity of aquatic ecosystems, such as disturbing aquatic habitats and exacerbating sediment disturbance. Against this backdrop, utilizing technologies such as geographic information systems, remote sensing monitoring, and intelligent algorithms to build modern fishery regulatory systems has become a common trend among countries to improve enforcement efficiency and reduce regulatory costs. These technologies are increasingly being applied in areas such as water boundary demarcation, vessel tracking, and risk area identification.

[0003] Currently, various technological solutions have been developed in the field of illegal fishing monitoring. For example, combining electronic navigation charts with the Global Positioning System (GPS) enables basic monitoring of vessel navigation areas; risk heat maps constructed using historical data on illegal activities can provide a reference for planning key enforcement areas; and some systems incorporate hydrological monitoring data, attempting to use parameters such as water flow and depth to assist in determining the legality of vessel operations. These technologies have played a role in their respective application scenarios. For instance, electronic fence technology can issue warnings to vessels entering specific areas, and multibeam sonar scanning data helps identify riverbed topographic features, providing a basis for distinguishing between legal operating areas and protected areas.

[0004] However, existing technologies still have limitations in dealing with complex and dynamic aquatic environments. On the one hand, most systems rely on static base map data, making it difficult to integrate real-time environmental parameters such as dynamic changes in water boundaries and sediment migration, resulting in insufficient accuracy in identifying dynamic shifts in protected area boundaries and spatiotemporal migrations of areas with high incidence of illegal activities. On the other hand, in vessel trajectory analysis, simple regional determinations are often made based solely on latitude and longitude coordinates, without establishing a suitable coordinate system that incorporates geographical features such as river curvature, making it difficult to accurately capture vessel evasion patterns. Summary of the Invention

[0005] The purpose of this invention is to provide an optimization method for illegal fishing based on base map features, addressing the following technical problems:

[0006] Existing monitoring systems rely on static base maps, making it difficult to integrate real-time parameters such as dynamic changes in water boundaries and sediment migration. This results in insufficient accuracy in identifying boundary shifts in protected areas and spatiotemporal migrations in high-incidence areas of illegal activity. Ship trajectory analysis is based solely on latitude and longitude for simple regional determinations, without establishing an adapted coordinate system that incorporates river curvature, making it difficult to accurately capture ship evasion of regulatory behaviors.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] An optimization method for illegal fishing based on base map features includes the following steps:

[0009] S1. Obtain a multi-source geographic information dataset of the target water area and generate a spatiotemporal continuous field of base map features, including the dynamic topological structure of the water area boundary, the distribution map of hydrological tomography parameters, and the heat map of historical illegal activities.

[0010] S2. Establish a manifold mapping coordinate system that is adaptive to the curvature of the river channel, and convert the latitude and longitude coordinates in the real-time ship trajectory data stream into a sequence of curvature arc length parameters in the manifold coordinate system.

[0011] S3. Calculate the normal penetration distance between the ship trajectory point and the dynamic topological protection zone unit of the water boundary in the manifold coordinate system, and at the same time detect the spatial overlap area between the trajectory line and the sediment boundary in the hydrographic parameter distribution map.

[0012] S4. Extract the distribution characteristics of the dwell time of the ship trajectory in the high-density area of ​​the heat map of historical illegal activities, and combine the real-time meteorological and hydrological data stream to correct the normal penetration distance threshold and the spatial overlap area weight.

[0013] S5. Generate an environmental adaptive risk field gradient map based on normal penetration distance, spatial overlap area, and dwell time, output multi-level law enforcement response instructions, and simultaneously optimize the spatial deployment scheme of monitoring equipment.

[0014] As a further aspect of the present invention: in S1, the generation process of the spatiotemporal continuum of the base map feature is as follows:

[0015] The dynamic topology of the water boundary specifically includes loading the basic data of the electronic navigation map of the water conservancy department and integrating the shoreline deformation trajectory function output by the flood inundation model to calculate the dynamic offset of the boundary of the protected area under different hydrological conditions;

[0016] The hydrographic parameter distribution map specifically includes a three-dimensional riverbed micro-geomorphic model based on multibeam sonar scanning data and extracts contour lines of sediment type boundaries and their migration rate curves as they are eroded by water flow.

[0017] The heat map of historical illegal activities specifically includes aggregating the set of geographic markers of historical illegal events and fusing point cloud data of ship loitering trajectories, and generating a time-slice heat distribution with standard deviation confidence intervals through Gaussian kernel density estimation.

[0018] As a further aspect of the present invention: in S2, the process of establishing a manifold mapping coordinate system that adapts to the river channel curvature is as follows:

[0019] First, identify the continuous curvature extrema of the river centerline, and divide the river into multiple differential arc segments using these extrema as nodes. Construct a local FRENERAL frame within each differential arc segment, and project the GPS coordinates of the ship trajectory points onto the tangential normal parameter space of the FRENERAL frame. Perform a first-order difference operation on the projected tangential parameter sequence to obtain the rate of change of manifold curvature of the ship trajectory.

[0020] As a further aspect of the present invention: in S3, the normal penetration distance is calculated by using the ray method to detect the positional relationship between the ship trajectory point and the polygon of the dynamic topological protection zone of the water boundary. When the trajectory point falls inside the polygon, a positive value of the penetration depth is recorded.

[0021] The detection of spatial overlap area is achieved through raster algebra operations, which involves performing a pixel-by-pixel logical AND operation between the ship trajectory buffer polygon and the sediment type raster of the hydrographic parameter distribution map.

[0022] As a further aspect of the present invention: In S4, the extraction of the distribution characteristics of the dwell time period adopts the sliding time window Fourier transform to analyze the spectral characteristics of the dwell period in the high-density area of ​​the heat map of historical illegal activities of the ship trajectory, and introduces the lunar phase cycle factor to modulate the distribution characteristics of the dwell time period. The lunar phase cycle factor is calculated by astronomical algorithm and has a value range of 0-1. It takes the minimum value during the new moon and the maximum value during the full moon. The dwell time period within the full moon cycle automatically increases the risk level.

[0023] The normal penetration distance threshold is dynamically adjusted based on real-time water turbidity monitoring values. When the real-time water turbidity monitoring values ​​increase, the normal penetration distance threshold range is automatically expanded. The spatial overlap area weight is adjusted in conjunction with real-time rainfall intensity data. Under heavy rainfall conditions, the decision weight of the spatial overlap area is reduced.

[0024] As a further aspect of the present invention: in step S5, the process of generating the environmental adaptive risk field gradient map is as follows:

[0025] The normalized value of the normal penetration distance, the weighted value of the spatial overlap area, and the risk level value of the dwell time period are integrated; the Laplacian operator convolution operation is performed on the integrated three-dimensional parameter space to extract the spatial distribution extreme value region of the risk gradient; the geometric center coordinates and risk intensity values ​​of the gradient extreme value region are marked to form a risk field topology map, which includes risk intensity contour lines and gradient direction arrow fields.

[0026] As a further aspect of the present invention: in step S5, the output of the three-level law enforcement response instruction specifically includes:

[0027] The multi-level response is divided into a first-level response, a second-level response, and a third-level response;

[0028] The triggering condition for a Level 1 response is that the normal penetration distance exceeds the correction threshold and the gradient strength of the risk field is greater than the critical value. At this time, the target coordinates and interception vector are sent to the law enforcement vessel.

[0029] The triggering condition for a Level II response is a continuous expansion of the spatial overlap area accompanied by a surge in the risk level during the dwell time. Under this condition, drones equipped with multispectral imagers are dispatched to collect evidence.

[0030] The criteria for determining a Level III response are based on the cross-regional risk field gradient linkage shown in the heat map of historical illegal activities. At this time, the collaborative response protocol of law enforcement units in adjacent waters is activated and risk field topology map data is shared.

[0031] As a further aspect of the present invention: in step S5, the process of optimizing the spatial deployment scheme of the monitoring equipment is as follows:

[0032] The spatial deployment scheme of the monitoring equipment includes a UAV patrol path planning module, which maps the gradient extreme value region of the risk field topology map into a UAV waypoint sequence;

[0033] The swirling scanning radius is allocated according to the gradient intensity value. A spiral progressive scanning mode is used in high-intensity areas to cover the core risk area. A pan-tilt control command set is generated for the fixed monitoring station so that the lens axis is continuously aligned with the risk field gradient transition interface at the boundary of multiple protection zones. The transition interface is defined by the dense area of ​​risk intensity contour lines.

[0034] As a further aspect of the present invention: when a ship is detected to be circumventing the monitoring, the trajectory characteristics of the ship circumventing the backwater area of ​​the water conservancy facility are analyzed, and the median radius of curvature and average speed of the circumvention path are extracted as path modeling parameters. Based on the path modeling parameters, a temporary monitoring enhancement zone is generated in the dynamic topology of the water boundary, and the early warning range of the sediment sensitive area in the hydrographic parameter distribution map is updated. The width of the enhancement zone is proportional to the radius of curvature.

[0035] The beneficial effects of this invention are:

[0036] This invention addresses the shortcomings of existing systems that rely on static base maps, such as the dynamic shift of protected area boundaries and insufficient accuracy in identifying the spatiotemporal migration of high-incidence illegal areas. It generates a spatiotemporal continuous field of base map features, including dynamic topological structures of water boundaries, hydrographic parameter distribution maps, and heat maps of historical illegal activity imprints. This achieves dynamic fusion and spatiotemporal correlation of multi-source geographic information. By establishing a manifold mapping coordinate system adaptive to river curvature, it converts latitude and longitude coordinates into a curvature arc length parameter sequence, overcoming the limitations of simple regional determination based solely on latitude and longitude. This improves the adaptability of vessel trajectory analysis to river geographic features and accurately captures the rate of change of trajectory curvature. Furthermore, it calculates... The algorithm corrects risk parameters by incorporating penetration distance, spatial overlap area, real-time meteorological and hydrological data, and lunar phase cycle factors, achieving environmentally adaptive risk assessment. This addresses the problem of insufficient correlation between existing models and environmental factors, improving the accuracy of risk level determination. By generating an environmentally adaptive risk field gradient map and outputting three-level enforcement response instructions, combined with UAV spiral scanning, dynamic alignment of fixed monitoring stations, and dynamic evolution mechanism of electronic fences, the spatial deployment of monitoring equipment is optimized, effectively responding to vessel evasion behavior and improving the targeting and collaborative efficiency of enforcement response. Ultimately, this achieves accurate identification, dynamic monitoring, and efficient handling of illegal fishing activities. Attached Figure Description

[0037] The invention will now be further described with reference to the accompanying drawings.

[0038] Figure 1 This is a flowchart illustrating the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0040] Please see Figure 1 As shown, this invention is an optimization method for illegal fishing based on base map features, comprising the following steps:

[0041] S1. Acquire multi-source geographic information datasets for the target water area. These datasets include basic data from electronic navigation charts provided by the water resources department, multibeam sonar scan data, and geographic marker sets of historical illegal activities. Based on these data, generate a spatiotemporal continuous field of base map features. This continuous field includes the dynamic topology of the water area boundaries, hydrographic parameter distribution maps, and heat maps of historical illegal activity imprints. Specifically, the dynamic topology of the water area boundaries integrates the shoreline deformation trajectory output by the flood inundation model; the hydrographic parameter distribution map reflects the three-dimensional riverbed micro-topography and sediment type boundaries; and the heat map of historical illegal activity imprints aggregates the density data of historical illegal events and vessel loitering trajectories.

[0042] S2. Establish a manifold mapping coordinate system that adapts to the river channel curvature. First, identify the continuous curvature extrema points of the river channel centerline. Using these points as nodes, divide the river channel into multiple differential arc segments. Construct a local Flyner frame within each arc segment. Then, project the latitude and longitude coordinates in the real-time ship trajectory data stream onto the tangential normal parameter space of the frame, converting it into a curvature arc length parameter sequence in the manifold coordinate system. Finally, obtain the rate of change of manifold curvature of the trajectory through first-order difference operations.

[0043] S3. Calculate the normal penetration distance between the ship trajectory point and the dynamic topological protection zone unit of the water boundary in the manifold coordinate system. Use the ray method to detect the positional relationship between the two. Record the positive value of the penetration depth when the trajectory point falls inside the polygon of the protection zone. At the same time, detect the spatial overlap area between the trajectory line and the sediment boundary in the hydrographic parameter distribution map. Perform a pixel-by-pixel logical AND operation between the trajectory buffer polygon and the sediment type raster through raster algebra operation.

[0044] S4. Extract the distribution characteristics of the dwell time period of the ship trajectory in the high-density area of ​​the heat map of historical illegal activities, and use the sliding time window Fourier transform to analyze the spectral characteristics of the dwell period; combine the real-time meteorological and hydrological data stream correction parameters, adjust the normal penetration distance threshold according to the real-time water turbidity, correlate the spatial overlap area weight with the real-time rainfall intensity, and introduce the lunar phase period factor to modulate the distribution characteristics of the dwell time period.

[0045] S5. Generate an environmentally adaptive risk field gradient map, integrating the normalized value of the normal penetration distance, the weighted value of the spatial overlap area, and the risk level value of the dwell time period. Extract the extreme value region of the risk gradient through Laplacian convolution operation, and mark the geometric center coordinates and intensity values ​​to form a topological map containing contour lines and gradient direction arrows. Based on this, output three-level law enforcement response instructions, simultaneously optimize the spatial deployment scheme of monitoring equipment, plan drone patrol paths, generate gimbal control instructions for fixed monitoring stations, and generate temporary monitoring enhancement zones when ship detour trajectories are detected.

[0046] In S1, the generation process of the spatiotemporal continuous field of the base map feature is as follows:

[0047] The construction process of the dynamic topology of the water boundary specifically includes loading basic data from the electronic navigation charts of the water resources department. This data covers basic geographic information such as channel depth, shoreline coordinates, and protected area demarcation. It also integrates the shoreline deformation trajectory function output by the flood inundation model. The flood inundation model comprehensively considers parameters such as basin rainfall, upstream inflow, and riverbed roughness to simulate water level changes under different hydrological conditions, thereby obtaining the shoreline deformation trajectory as the water level rises and falls. Based on this data, by calculating the dynamic offset of the protected area boundary relative to the baseline shoreline under different hydrological conditions, a topology that reflects the real-time changes in the boundary is formed. This topology includes both the stable boundary during the dry season and the temporary inundation boundary during the flood season.

[0048] The generation process of the hydrotomographic parameter distribution map specifically includes analyzing multibeam sonar scanning data. Multibeam sonar transmits multiple beams of sound waves underwater and receives reflected signals to construct a three-dimensional micro-topographic model of the riverbed. This model contains detailed topographic information such as riverbed undulations, shoal distribution, and reef locations. Sediment type boundary contour lines are extracted from this model. Based on the acoustic reflection characteristics of sediments, different types such as silt, gravel, and rock are distinguished. Simultaneously, combined with water flow monitoring data, the migration patterns of various sediment types under water scouring are analyzed, generating migration rate curves to reflect the movement trends of sediment boundaries under different seasons and flow conditions. Finally, a dynamically updated hydrotomographic parameter distribution map is generated.

[0049] The generation process of the heat map of historical illegal activities includes aggregating a set of geographic markers for historical illegal events. These markers come from data such as the location of illegal fishing sites and illegal vessel mooring points recorded in law enforcement records. Simultaneously, it integrates point cloud data of vessel loitering trajectories. This data is collected through vessel BeiDou or GPS positioning systems, recording the vessel's stopping and circling trajectories in specific areas. Gaussian kernel density estimation is used to process this data, transforming the discrete point data into a continuous density distribution, generating a time-slice heat map with standard deviation confidence intervals. Each time slice corresponds to the density of illegal activities during a specific period. By overlaying consecutive slices, the spatiotemporal evolution characteristics of illegal activities can be visually presented.

[0050] The specific steps for generating the spatiotemporal continuum of the base map features are as follows:

[0051] Establish a unified geographic reference coordinate system and time axis benchmark. The geographic reference coordinate system adopts the national unified geodetic coordinate system to ensure the accurate alignment of the three types of data in spatial location. The time axis benchmark uses standard timestamps as the unit to unify the time recording standard of various types of data. The spatial resolution of the three types of elements is resampled to the same raster size to keep the data from different sources consistent in spatial scale, which is convenient for subsequent overlay analysis.

[0052] Interpolating the shoreline deformation trajectory function of the dynamic topology of the water boundary in the time dimension, and filling the boundary data gaps between different monitoring times through interpolation operations, generates a continuous time series boundary displacement field that can reflect the boundary position at any time in real time.

[0053] The sediment migration rate curves from the hydrotomographic parameter distribution map are converted into a spatiotemporal diffusion model. The model combines parameters such as water flow direction and velocity to predict the distribution of sediment parameters at different future time periods, so that the distribution map not only includes the current state but also reflects future trends.

[0054] The time-slice heat distribution of the heat map of historical illegal activities is processed by a temporal convolutional neural network. The neural network learns the correlation between different time slices, expanding the discrete time slices into a continuous heat field, thereby enabling the estimation of the risk intensity of historical illegal activities at any given time.

[0055] Finally, the boundary displacement field, parameter prediction field and thermodynamic field are superimposed to form a four-dimensional spatiotemporal continuous field, in which each spatial coordinate point is associated with the dynamic boundary state, the corresponding sediment parameter value and the historical risk intensity value. Through this continuous field, comprehensive geographical and risk information of any spatiotemporal point can be intuitively obtained.

[0056] In S2, the process of establishing a manifold mapping coordinate system that adapts to the curvature of the river channel is as follows:

[0057] First, the centerline of the target river channel is refined and extracted. A continuous centerline curve is obtained by fitting the river's geometry. Based on this, continuous curvature extrema of the curve are identified. Curvature extrema refer to the locations where the curve's curvature reaches a local maximum or minimum. These points reflect significant changes in the river's direction, such as convex or concave points at sharp bends. Using these curvature extrema as nodes, the entire river channel centerline is divided into multiple differential arc segments. The length of each differential arc segment is dynamically adjusted according to the severity of the river's curvature change. Arc segments can be appropriately lengthened in areas of gentle curvature change and shortened in areas of dramatic curvature change, ensuring that the geometry within each arc segment approximates a regular curve.

[0058] Within each differential arc segment, a local Flexner frame is constructed. This frame is a dynamically changing local coordinate system that varies with the arc segment's orientation, consisting of a tangent vector, a normal vector, and a binormal vector. The tangent vector points along the tangent direction of the arc segment towards the river's extension direction; the normal vector is perpendicular to the tangent vector and points towards the inside or outside of the river channel; and the binormal vector is perpendicular to the vertical direction formed by the tangent and normal vectors in the three-dimensional space. The GPS coordinates (latitude and longitude) of the ship's trajectory points are projected onto the tangential and normal parameter space of this Flexner frame using a coordinate transformation algorithm. During the transformation, the effects of Earth's curvature and projection distortion must be eliminated, ensuring that the trajectory points are represented in the local coordinate system by tangential and normal distances.

[0059] The first-order difference operation is performed on the tangential parameter sequence obtained after projection, that is, the difference between the tangential parameters of two adjacent trajectory points is calculated. This difference can reflect the rate of change of the ship's movement along the direction of the river channel, and then be converted into the rate of change of the manifold curvature of the ship's trajectory. This quantifies the degree to which the ship's navigation trajectory adapts to the river channel's curvature. For example, when the rate of change of trajectory curvature is large, it indicates that the ship's turning action is violent and there may be abnormal navigation behavior.

[0060] In step S3, the normal penetration distance is calculated using the ray-mapping method to detect the positional relationship between the ship's trajectory point and the polygon of the dynamic topological protection zone of the water boundary. The ray-mapping method involves emitting a ray from the ship's trajectory point in a predetermined direction, typically perpendicular to the normal of the river centerline. The intersection of this ray with the boundary of the protection zone polygon is then detected. When the trajectory point is inside the protection zone polygon, the distance from the trajectory point to the boundary intersection is recorded as the penetration depth and marked as a positive value; a larger value indicates a deeper penetration into the protection zone. If the trajectory point is outside the protection zone, and the ray does not intersect the boundary or the intersection point is outside the trajectory point, a value of zero or negative is recorded to distinguish different spatial relationships.

[0061] The detection of spatial overlap area is achieved through raster algebra operations. First, the area containing the ship's trajectory is rasterized. An appropriate raster size is set according to the monitoring accuracy requirements, dividing the continuous spatial area into a large number of regular raster units. A buffer polygon is constructed for the ship's trajectory. The width of the buffer is determined based on the ship's size and possible operating range, expanding outwards from the trajectory line to form a polygonal region. Simultaneously, sediment type data from the hydrographic parameter distribution map is converted into a raster format, with each raster unit corresponding to a specific sediment type, such as silt, sand, or rock. Through pixel-by-pixel logical AND operations, the raster of the ship trajectory buffer polygon is compared with the sediment type raster to determine whether each raster unit simultaneously belongs to both the buffer zone and the target sediment type area. The number of all raster units meeting the conditions is counted, and the total spatial overlap area is calculated based on the area of ​​each individual raster, thus reflecting the spatial correlation between the ship's trajectory and a specific sediment region.

[0062] In step S4, the extraction of the distribution characteristics of dwell time periods employs a sliding time window Fourier transform. The sliding time window moves segment by segment along the time axis, using fixed duration units, covering all dwell records within the high-density areas of the historical illegal activity heatmap. Each time window contains data such as dwell duration and frequency for that period. The data within each time window is processed using Fourier transform, converting the dwell characteristics in the time domain into a periodic spectrum in the frequency domain. This allows for analysis of the periodic patterns of dwell behavior, such as the existence of fixed daily dwell time periods or weekly dwell peaks, thereby identifying time characteristics consistent with illegal fishing patterns.

[0063] A lunar phase cycle factor is introduced to modulate the distribution characteristics of dwell time periods. This factor is calculated using an astronomical algorithm based on the Moon's orbital parameters around the Earth, combined with the relative positions of the Sun, Earth, and Moon, to determine the lunar phase at any given time. Its value ranges from 0 to 1; during a new moon, the Moon is between the Earth and the Sun, resulting in the weakest ground illumination, and the factor has its minimum value; during a full moon, the Earth is between the Moon and the Sun, resulting in the strongest ground illumination, and the factor has its maximum value. This modulation synchronizes the risk assessment of dwell time periods with lunar phase changes. During a full moon cycle, dwell time periods of vessels in high-density areas are automatically assigned a higher risk level to accommodate the potential impact of moonlight conditions on illegal nighttime fishing activities.

[0064] The normal penetration distance threshold is dynamically adjusted based on real-time turbidity monitoring values. Real-time turbidity reflects the concentration of suspended particles in the water. Increased turbidity reduces visibility, making it more difficult to determine the actual operating range of vessels. In this case, the normal penetration distance threshold range is automatically expanded, making the system's criteria for determining whether a vessel is entering the protected area more lenient and preventing missed detections due to deteriorating monitoring conditions.

[0065] The weighting of spatial overlap area is adjusted in relation to real-time rainfall intensity data. During heavy rainfall, increased water flow velocity and intensified sediment disturbance reduce the stability of sediment boundaries in hydrographic parameter distribution maps, potentially leading to deviations in the calculation of the spatial overlap area between ship trajectories and sediment boundaries. Therefore, reducing the decision weight of spatial overlap area under heavy rainfall conditions minimizes the impact of meteorological interference on risk assessment and ensures that the assessment results more closely reflect reality.

[0066] In step S5, the process of generating the environmental adaptive risk field gradient map is as follows:

[0067] First, the normalized value of the normal penetration distance, the weighted value of the spatial overlap area, and the risk level value of the dwelling time period are integrated. The normalized value of the normal penetration distance converts the actual penetration distance into a relative value between 0 and 1, eliminating the influence of different water scales on the distance parameter; the weighted value of the spatial overlap area is a comprehensive value calculated by combining the corrected weights, reflecting the actual contribution of the overlap area in risk assessment; the risk level value of the dwelling time period is the final risk score after lunar phase modulation, reflecting the likelihood of violation in the time dimension.

[0068] The integrated three-dimensional parameter space is subjected to Laplace convolution. The Laplace operator captures the rate of change of parameter values ​​by calculating the second derivative at each point in the space, thereby extracting the extreme regions of the spatial distribution of risk gradient. These regions are where the risk values ​​change most drastically, and may be key areas for the transition of illegal activities from low-risk to high-risk areas, or concentrated areas of high-risk behavior.

[0069] The geometric center coordinates and risk intensity values ​​of gradient extrema regions are marked to form a risk field topology map. Risk intensity contour lines in the map connect points with the same risk value, visually demonstrating the spatial distribution and severity differences of risk. Gradient direction arrows indicate the direction of risk value change from low to high, clearly showing the trend of risk diffusion or aggregation. This map comprehensively reflects the distribution characteristics of illegal fishing risks in target waters, providing precise data for subsequent law enforcement responses and monitoring deployments.

[0070] In S5, the output of the three-level law enforcement response instruction specifically includes:

[0071] The multi-level response is divided into Level 1, Level 2, and Level 3 responses, forming a hierarchical and progressive law enforcement dispatch mechanism, among which:

[0072] The trigger condition for a Level 1 response is that the normal penetration distance exceeds a correction threshold and the risk field gradient intensity is greater than a critical value. When the normal penetration distance of a vessel's trajectory point entering the protected area exceeds the dynamically adjusted threshold, and the risk gradient intensity at that location reaches a preset critical value, the system automatically generates a Level 1 response command. At this time, the system sends the real-time coordinates and interception vector of the target vessel to the nearest law enforcement vessel. The interception vector comprehensively considers factors such as the target vessel's heading, speed, and water topography to calculate the optimal interception path, guiding the law enforcement vessel to reach the target location in the shortest possible time for interception.

[0073] The trigger condition for a Level II response is a continuous expansion of the spatial overlap area accompanied by a surge in the risk level during the vessel's stay. A Level II response is triggered when the spatial overlap area between the vessel's trajectory and sensitive sediment areas on the hydrographic parameter distribution map continuously increases, and simultaneously, the risk level of the vessel's stay in a high-density area on the heat map of historical illegal activity imprints significantly increases due to factors such as lunar phase modulation. Under this condition, a drone equipped with a multispectral imager is deployed, flies along a pre-set route to the target area, and uses different bands of the multispectral imager to capture images of the vessel and its surrounding environment from multiple angles, collecting evidence such as vessel operating equipment and fishing gear, providing intuitive image data for subsequent law enforcement.

[0074] The determination of a Level III response is based on the cross-regional risk field gradient linkage displayed by the heat map of historical illegal activities. When the extreme value areas of the risk field gradient in different waters show a synchronous changing trend, indicating the possible existence of organized illegal fishing activities across regions, the collaborative response protocol of law enforcement units in adjacent waters is activated. At this time, risk field topology map data is shared, including information such as risk intensity contour lines and gradient direction arrow fields, enabling each law enforcement unit to fully grasp the risk distribution situation within the region, uniformly allocate law enforcement resources, form a joint law enforcement force, and carry out a comprehensive crackdown on cross-regional illegal activities.

[0075] In step S5, the process of optimizing the spatial deployment scheme of monitoring equipment is as follows:

[0076] The spatial deployment scheme for monitoring equipment includes a drone patrol path planning module. This module maps the gradient extreme value regions of the risk field topology map into a sequence of drone waypoints. Gradient extreme value regions are areas of most dramatic risk changes, typically corresponding to areas with high incidence of illegal activities. Converting the center coordinates of these regions into drone patrol waypoints ensures that drones can focus on monitoring high-risk areas.

[0077] The circling scanning radius is allocated based on the gradient intensity value. The higher the gradient intensity value, the more concentrated the risk in the area and the greater the possibility of illegal activities. For high-intensity areas, a spiral progressive scanning mode is adopted, with the drone gradually descending from the periphery towards the center to expand the scanning range and cover the core risk area, ensuring comprehensive monitoring of high-risk targets.

[0078] A set of pan-tilt-zoom (PTZ) control commands is generated for the fixed monitoring station to ensure that the camera axis is continuously aligned with the risk field gradient transition interface at the boundary of multiple protected areas. The risk field gradient transition interface is defined by densely packed risk intensity contour lines; these areas exhibit significant changes in risk level and represent potential pathways for illegal activities to shift from one protected area to another. By adjusting the PTZ angle and focal length of the fixed monitoring station, the monitoring camera is kept consistently focused on these key interfaces, enabling real-time capture of vessel cross-regional movement and improving monitoring capabilities in the boundary areas.

[0079] In a preferred embodiment of the present invention, when a vessel is detected to be circumventing monitoring, the focus is on analyzing the vessel's navigation characteristics in the backwater area of ​​the hydraulic facility. The backwater area of ​​a hydraulic facility typically has gentle currents, easily creating monitoring blind spots. Vessels often use this area to change course and avoid the regular monitoring range. By tracing the vessel's complete trajectory in this area, key parameters of the evasion path are extracted. The median radius of curvature reflects the typical turning amplitude of the vessel, and the average speed reflects the vessel's rhythm during the evasion process; both together constitute the core basis for path modeling.

[0080] Based on these path modeling parameters, temporary monitoring enhancement zones are generated within the dynamic topology of the water boundary. These enhancement zones are deployed along the outer edge of the vessel avoidance path, covering potential extended routes. Their morphology adjusts in real-time with the dynamic changes of the water boundary to ensure consistency with the topological relationship of the protected area boundary. Simultaneously, considering the potential impact of vessel detours on the surrounding hydrological environment, the early warning range for sediment-sensitive areas on the hydrotomographic parameter distribution map is updated, incorporating areas potentially disturbed by vessel propeller turbulence into the early warning system.

[0081] The width of the temporary monitoring enhancement zone is proportional to the radius of curvature of the vessel's avoidance path; that is, the greater the vessel's turning radius, the wider the coverage area of ​​the enhancement zone, to match the vessel's possible operating radius. This dynamic adjustment mechanism allows monitoring resources to quickly focus on areas with high incidence of vessel avoidance behavior, avoiding resource waste caused by over-monitoring while ensuring effective tracking of potential illegal activities. It complements the conventional monitoring network, enhancing the overall flexibility and targeting of supervision.

[0082] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A method for optimizing illegal fishing based on base map features, characterized in that, Includes the following steps: S1. Obtain a multi-source geographic information dataset of the target water area and generate a spatiotemporal continuous field of base map features, including the dynamic topological structure of the water area boundary, the distribution map of hydrological tomography parameters, and the heat map of historical illegal activities. S2. Establish a manifold mapping coordinate system that is adaptive to the curvature of the river channel, and convert the latitude and longitude coordinates in the real-time ship trajectory data stream into a sequence of curvature arc length parameters in the manifold coordinate system. S3. Calculate the normal penetration distance between the ship trajectory point and the dynamic topological protection zone unit of the water boundary in the manifold coordinate system, and at the same time detect the spatial overlap area between the trajectory line and the sediment boundary in the hydrographic parameter distribution map. S4. Extract the distribution characteristics of the dwell time of the ship trajectory in the high-density area of ​​the heat map of historical illegal activities, and combine the real-time meteorological and hydrological data stream to correct the normal penetration distance threshold and the spatial overlap area weight. S5. Generate an environmental adaptive risk field gradient map based on normal penetration distance, spatial overlap area, and dwell time, output multi-level law enforcement response instructions, and simultaneously optimize the spatial deployment scheme of monitoring equipment.

2. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In S1, the generation process of the spatiotemporal continuous field of the base map feature is as follows: The dynamic topology of the water boundary specifically includes loading the basic data of the electronic navigation map of the water conservancy department and integrating the shoreline deformation trajectory function output by the flood inundation model to calculate the dynamic offset of the boundary of the protected area under different hydrological conditions; The hydrographic parameter distribution map specifically includes a three-dimensional riverbed micro-geomorphic model based on multibeam sonar scanning data and extracts contour lines of sediment type boundaries and their migration rate curves as they are eroded by water flow. The heat map of historical illegal activities specifically includes aggregating the set of geographic markers of historical illegal events and fusing point cloud data of ship loitering trajectories, and generating a time-slice heat distribution with standard deviation confidence intervals through Gaussian kernel density estimation.

3. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In S2, the process of establishing a manifold mapping coordinate system that adapts to the curvature of the river channel is as follows: First, identify the continuous curvature extrema of the river centerline, and divide the river into multiple differential arc segments using these extrema as nodes. Construct a local FRENERAL frame within each differential arc segment, and project the GPS coordinates of the ship trajectory points onto the tangential normal parameter space of the FRENERAL frame. Perform a first-order difference operation on the projected tangential parameter sequence to obtain the rate of change of manifold curvature of the ship trajectory.

4. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In S3, the normal penetration distance is calculated by using the ray method to detect the positional relationship between the ship trajectory point and the polygon of the dynamic topological protection zone of the water boundary. When the trajectory point falls inside the polygon, the penetration depth is recorded as a positive value. The detection of spatial overlap area is achieved through raster algebra operations, which involves performing a pixel-by-pixel logical AND operation between the ship trajectory buffer polygon and the sediment type raster of the hydrographic parameter distribution map.

5. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In S4, the extraction of the distribution characteristics of the stay time period adopts the sliding time window Fourier transform to analyze the spectral characteristics of the stay period in the high-density area of ​​the heat map of historical illegal activities. The lunar phase cycle factor is introduced to modulate the phase of the stay time period distribution characteristics. The lunar phase cycle factor is calculated by astronomical algorithm and has a value range of 0-1. It takes the minimum value during the new moon and the maximum value during the full moon. The stay time period within the full moon cycle automatically increases the risk level. The normal penetration distance threshold is dynamically adjusted based on real-time water turbidity monitoring values. When the real-time water turbidity monitoring values ​​increase, the normal penetration distance threshold range is automatically expanded. The spatial overlap area weight is adjusted in conjunction with real-time rainfall intensity data. Under heavy rainfall conditions, the decision weight of the spatial overlap area is reduced.

6. The method for optimizing illegal fishing based on base map features according to claim 5, characterized in that, In step S5, the process of generating the environmental adaptive risk field gradient map is as follows: The normalized value of the normal penetration distance, the weighted value of the spatial overlap area, and the risk level value of the dwell time period are integrated; the Laplacian operator convolution operation is performed on the integrated three-dimensional parameter space to extract the spatial distribution extreme value region of the risk gradient; the geometric center coordinates and risk intensity values ​​of the gradient extreme value region are marked to form a risk field topology map, which includes risk intensity contour lines and gradient direction arrow fields.

7. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In S5, the output of the three-level law enforcement response instruction specifically includes: The multi-level response is divided into a first-level response, a second-level response, and a third-level response; The triggering condition for a Level 1 response is that the normal penetration distance exceeds the correction threshold and the gradient strength of the risk field is greater than the critical value. At this time, the target coordinates and interception vector are sent to the law enforcement vessel. The triggering condition for a Level II response is a continuous expansion of the spatial overlap area accompanied by a surge in the risk level during the dwell time. Under this condition, drones equipped with multispectral imagers are dispatched to collect evidence. The criteria for determining a Level III response are based on the cross-regional risk field gradient linkage shown in the heat map of historical illegal activities. At this time, the collaborative response protocol of law enforcement units in adjacent waters is activated and risk field topology map data is shared.

8. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, In step S5, the process of optimizing the spatial deployment scheme of monitoring equipment is as follows: The spatial deployment scheme of the monitoring equipment includes a UAV patrol path planning module, which maps the gradient extreme value region of the risk field topology map into a UAV waypoint sequence; The swirling scanning radius is allocated according to the gradient intensity value. A spiral progressive scanning mode is used in high-intensity areas to cover the core risk area. A pan-tilt control command set is generated for the fixed monitoring station so that the lens axis is continuously aligned with the risk field gradient transition interface at the boundary of multiple protection zones. The transition interface is defined by the dense area of ​​risk intensity contour lines.

9. The method for optimizing illegal fishing based on base map features according to claim 1, characterized in that, When a vessel is detected to be circumventing monitoring, the trajectory characteristics of the vessel circumventing the backwater area of ​​the water conservancy facility are analyzed. The median radius of curvature and the average speed of movement of the circumvention path are extracted as path modeling parameters. Based on the path modeling parameters, a temporary monitoring enhancement zone is generated in the dynamic topology of the water boundary. At the same time, the warning range of sediment sensitive areas in the hydrographic parameter distribution map is updated. The width of the enhancement zone is proportional to the radius of curvature.

Images