Multi-source spatio-temporal data fusion two-dimensional and three-dimensional rendering and target automatic focusing method and system

Abstract

The application discloses a kind of based on multi-source space-time data fusion two three-dimensional unified rendering and key target automatic focusing method and system, it is related to geographic information system technical field.The method includes: obtaining at least two kinds of target related space-time data, executing coordinate system, time reference unification and data reliability calibration;According to spatial proximity, time proximity, target type consistency and data source reliability, calculate correlation score, fuse data from different data sources into the same target and construct unified trajectory object;According to target attribute and state, determine target importance, according to this, determine aggregation and layered rendering strategy;According to trajectory reliability and map state, determine rendering level, map unified trajectory object is respectively mapped to two-dimensional and three-dimensional map and executed unified rendering;According to target importance, field of vision correlation degree and alarm state, determine focusing target and adjust map field of vision;When map perspective changes, update each state based on unified trajectory object.

Description

Technical Field

[0001] This invention relates to the field of geographic information system technology, and in particular to a method and system for multi-source spatiotemporal data fusion, 2D / 3D rendering, and automatic target focusing. Background Technology

[0002] In the fields of Geographic Information Systems (GIS) and Situational Awareness, existing map situational awareness systems typically integrate data from positioning devices, satellite positioning, video recognition, business events, alarms, and map regions, displaying target locations, historical trajectories, real-time trajectories, and alarm trajectories on 2D or 3D maps. Furthermore, existing systems load, thin, or aggregate trajectory points based on map zoom levels, screen field of view, or data volume thresholds to handle the rendering pressure from large-scale target data. However, these methods still have many shortcomings in scenarios involving multiple targets, large areas, and simultaneous 2D / 3D display.

[0003] Specifically, the coordinate systems, timestamps, sampling periods, and data reliability of target data from different sources are often inconsistent. The same target is prone to forming multiple discontinuous or inconsistent trajectory representations, affecting subsequent rendering, playback, and focusing processing. Two-dimensional map layers and three-dimensional map entities usually use different data structures and rendering logics. Historical trajectories, real-time trajectories, and alarm trajectories are difficult to maintain the consistency of the same target's state when switching between two-dimensional and three-dimensional modes. When multiple targets are displayed simultaneously, ordinary target trajectories, key target trajectories, and alarm information of abnormal targets are prone to mutual occlusion. Existing aggregation displays usually only process based on spatial distance or zoom level, without incorporating target importance, abnormal status, and user attention behavior into the aggregation rules. In addition, existing systems usually rely on manual retrieval or manual adjustment of the viewpoint to locate key targets, lacking an automatic focusing loop based on target importance, alarm status, and user lock. When the map is translated, zoomed, rotated, or switched between two-dimensional and three-dimensional modes, the trajectory loading status, aggregation cluster expansion status, and focusing status lack a unified linkage mechanism, resulting in inconsistent display results. Summary of the Invention

[0004] To address the aforementioned technical problems in related technologies, this invention proposes a method and system for multi-source spatiotemporal data fusion 2D / 3D rendering and automatic target focusing, which can overcome the above-mentioned shortcomings of existing technologies.

[0005] To achieve the above-mentioned technical objectives, the technical solution of the present invention is implemented as follows: A method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocus; This multi-source spatiotemporal data fusion method for 2D / 3D rendering and automatic target focusing includes: Acquire at least two types of target-related spatiotemporal data; The acquired target-related spatiotemporal data are subjected to coordinate system one, time reference unification, and data reliability calibration. A correlation score is calculated based on spatial proximity, temporal proximity, target type consistency, and data source credibility. Data from different data sources are then merged into data for the same target based on the correlation score. A unified trajectory object is constructed based on the fused target data. The unified trajectory object includes at least the target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The importance of a target is determined based on its attributes, abnormal states, trajectory reliability, regional importance, and user interaction behavior. The aggregation score is calculated based on spatial distance, temporal similarity, direction of motion, target importance, and abnormal state, and the multi-target aggregation strategy and layered rendering strategy are determined based on the aggregation score. The rendering level of a trajectory segment is determined based on trajectory reliability, map status, and target importance. The same unified trajectory object is mapped to a two-dimensional map layer and a three-dimensional map entity respectively, and a unified two-dimensional and three-dimensional rendering is performed. The current focus target is determined based on the target's importance, its relevance to the current field of view, alarm status, and user lock status, and the map field of view parameters are automatically adjusted according to the current display mode. When the map is panned, zoomed, rotated, or switched between 2D and 3D, the trajectory loading status, aggregation status, and focus status are updated based on the unified trajectory object.

[0006] Furthermore, the target-related spatiotemporal data includes at least two of the following: positioning device data, satellite positioning data, mobile terminal location data, target location data output by the video recognition system, service event data, alarm data, map area data, and user interaction behavior data; the acquired individual data entries are organized into a unified access record R: R = {source_id, target_id, time, lon, lat, x, y, z, speed, heading, type, status, alarm, confidence}, Wherein, source_id represents the data source identifier, target_id represents the target identifier, time represents the data timestamp, lon and lat represent latitude and longitude, x, y, and z represent spatial coordinates, speed represents speed, heading represents direction, type represents target type, status represents behavior status, alarm represents alarm status, and confidence represents data source confidence.

[0007] Furthermore, implementing coordinate system one includes: transforming latitude and longitude coordinates, projected coordinates, and three-dimensional coordinates from different sources to a unified spatial coordinate system. The two-dimensional coordinate transformation relationship is as follows: P_2d = T_geo(lon, lat, epsg_source, epsg_target), The three-dimensional coordinate transformation relationship is as follows: P_3d = {x, y, h + h_offset}, Where h_offset is a three-dimensional height correction parameter, which is configured based on map elevation benchmark, device height benchmark, or business scenario. The execution of time benchmark unification includes: time alignment of data from different sources according to a preset time granularity delta_t. t' = floor(t / delta_t) × delta_t, In the real-time display scenario, delta_t is configured to be between 1 and 5 seconds, while in the historical playback scenario, delta_t is configured to be between 5 and 60 seconds.

[0008] Furthermore, the association score M is calculated as follows: M = a1×S_space + a2×S_time + a3×S_type + a4×S_source, a1 + a2 + a3 + a4 = 1, Where S_space represents spatial proximity, S_time represents temporal proximity, S_type represents target type consistency, S_source represents data source credibility, and a1, a2, a3 and a4 are the corresponding weights. When M is greater than the association threshold theta_match, the corresponding data will be merged into the same target. When M is less than or equal to theta_match, a new target candidate object will be generated or it will enter the manual confirmation queue. The associated threshold theta_match ranges from 0 to 1, with a1=0.40, a2=0.25, a3=0.20, and a4=0.15.

[0009] Furthermore, the unified trajectory object T is represented as: T = {target_id, time_series, point_2d_series, point_3d_series, track_type, confidence_series, importance_score, render_level, focus_state}, Among them, time_series represents time series, point_2d_series represents two-dimensional trajectory point series, point_3d_series represents three-dimensional trajectory point series, track_type represents historical trajectory, real-time trajectory, predicted trajectory or alarm trajectory, confidence_series represents trajectory confidence series, importance_score represents target importance, render_level represents rendering level, and focus_state represents focus state. The trajectory is divided into multiple trajectory segments according to a fixed time window, fixed spatial length, direction change point, or scene boundary. The confidence level C(i,k) is calculated for each trajectory segment. C(i,k) = λ1×S_source(i,k) + λ2×S_continuity(i,k) + λ3×S_motion(i,k) + λ4×S_constraint(i,k), λ1+λ2+λ3+λ4=1, Where S_source represents the reliability of the data source, S_continuity represents the continuity of the trajectory, S_motion represents the kinematic consistency, S_constraint represents the satisfaction of the scene constraints, and λ1=0.30, λ2=0.25, λ3=0.25, λ4=0.20; The reliability of the data source is determined based on the data source level, sampling completeness, and time delay; the trajectory continuity is determined based on the time interval and spatial interval between adjacent trajectory segments; the kinematic consistency is determined based on changes in speed, direction, and altitude; and the scene constraint satisfaction is determined based on constraints such as roads, area boundaries, restricted areas, or business areas.

[0010] Furthermore, the target importance I is calculated as follows: I = b1×S_type + b2×S_alarm + b3×S_confidence + b4×S_user + b5×S_region, b1+b2+b3+b4+b5=1, Where S_type represents the target type score, S_alarm represents the anomaly or alarm score, S_confidence represents the trajectory confidence score, S_user represents the user attention behavior score, S_region represents the importance score of the target's region, and b1=0.20, b2=0.30, b3=0.20, b4=0.15, b5=0.15; When I is greater than the first threshold theta_focus, the target is added to the priority target queue; when I is greater than the second threshold theta_alarm, the target is added to the forced focus queue; theta_focus ranges from 0 to 1, theta_alarm ranges from 0 to 1, and theta_alarm is greater than theta_focus.

[0011] Furthermore, the aggregation score G is calculated as follows: G = c1×D_space + c2×D_time + c3×D_heading + c4×(1 - I_norm) + c5×S_normal, c1+c2+c3+c4+c5=1, Wherein, D_space represents the normalized value of spatial distance, D_time represents the normalized value of time difference, D_heading represents the normalized value of orientation difference, I_norm represents the normalized value of target importance, S_normal represents the score of non-abnormal state, and c1=0.30, c2=0.20, c3=0.15, c4=0.25, c5=0.10; Targets with higher importance than theta_focus, abnormal targets, and user-locked targets are not included in normal aggregation; The layered rendering strategy outputs layered rendering results according to the following layers: base layer, common target aggregation layer, common target trajectory layer, key target trajectory layer, abnormal target alarm layer, and user's current focus target layer.

[0012] Furthermore, the rendering level L(i,k) is determined in the following manner: L(i,k) = H(C(i,k), Z, V(i,k), B, I), Where C(i,k) represents the reliability of the trajectory segment, Z represents the map zoom level, V(i,k) represents the positional relationship between the trajectory segment and the current field of view, B represents the current rendering resource status, and I represents the importance of the target; The rendering levels include at least: a first rendering level, which retains the trajectory index and confidence summary but does not generate complete graphic data; a second rendering level, which loads trajectory points at a sampling ratio of 10% to 50% and generates graphic data with an update cycle of 1 to 5 seconds; and a third rendering level, which loads complete trajectory points and generates graphic data with an update cycle of 0.1 to 1 second.

[0013] Furthermore, the mapping relationship for the unified 2D / 3D rendering is as follows: Render(T, mode) = {Render2D(T.point_2d_series, T.render_level),Render3D(T.point_3d_series, T.render_level)}, Here, mode represents the two-dimensional display mode, the three-dimensional display mode, or the two-dimensional and three-dimensional linkage display mode; two-dimensional rendering uses point_2d_series to generate trajectory lines, trajectory points, target icons, aggregate icons, and alarm markers; three-dimensional rendering uses point_3d_series to generate three-dimensional trajectory lines, three-dimensional target entities, height lines, dynamic trails, and timeline playback objects. The target selection score F is calculated as follows: F = d1×I + d2×S_visible + d3×S_alarm + d4×S_user_lock, When the forced focus queue is not empty, the target with the highest score is selected as the current focus target; when the forced focus queue is empty, the target with the highest overall score is selected as the candidate focus target from the priority target queue; when the user manually locks a target, the user-locked target remains the current focus target until the user unlocks the target or the target becomes invalid. The map view parameters include the map center point, 2D map zoom level, 3D view height, 3D pitch angle, target display level, label display status, and trajectory playback time window.

[0014] According to another aspect of the present invention, a multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocus system is provided; This multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocus system includes: The data access module is used to acquire at least two types of target-related spatiotemporal data; The spatiotemporal unification module is used to perform coordinate system unification, time reference unification, and data reliability calibration on the acquired target-related spatiotemporal data; The target association and fusion module is used to calculate the association score based on spatial proximity, temporal proximity, target type consistency and data source credibility, and to fuse data from different data sources into data of the same target based on the association score; A unified trajectory object construction module is used to construct a unified trajectory object based on the fused target data. The unified trajectory object includes at least target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The target importance assessment module is used to determine the importance of a target based on its attributes, abnormal status, trajectory reliability, regional importance, and user interaction behavior. The aggregation and rendering strategy decision module is used to calculate the aggregation score based on spatial distance, temporal similarity, motion direction, target importance and abnormal state, and determine the multi-target aggregation strategy and layered rendering strategy based on the aggregation score; The rendering level decision module is used to determine the rendering level of a trajectory segment based on trajectory credibility, map status, and target importance. The unified 2D / 3D rendering module is used to map the same unified trajectory object to a 2D map layer and a 3D map entity respectively, and perform unified 2D / 3D rendering. The autofocus module is used to determine the current focus target based on the target's importance, the correlation between the target and the current field of view, alarm status, and user lock status, and automatically adjust the map field of view parameters according to the current display mode. The status linkage update module is used to update the trajectory loading status, aggregation status, and focus status based on the unified trajectory object when the map is translated, zoomed, rotated, or switched between 2D and 3D.

[0015] The beneficial effects of this invention are as follows: By unifying the coordinate system, time base, and data reliability calibration, target data from different sources are integrated into a unified data foundation, reducing the problems of multiple trajectories and inconsistent states for the same target; by encapsulating two-dimensional and three-dimensional trajectory point sequences and rendering states through a unified trajectory object, two-dimensional and three-dimensional maps can be displayed and updated in conjunction with the same data structure; by using trajectory reliability and target importance to participate in rendering level decisions, data loading and graphics updates can be dynamically adjusted according to the target state and map state; by excluding key targets and abnormal targets from ordinary aggregation and setting a layered rendering order, the occlusion of key information by ordinary targets is reduced; by calculating focus scores based on target importance and alarm status and automatically adjusting map view parameters, the amount of manual retrieval and manual adjustment of the viewpoint is reduced; by updating each state based on the unified trajectory object when switching maps and changing the viewpoint, the continuity and consistency of the display results are ensured. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 : Overall method flowchart of the present invention; Figure 2 : System module structure diagram of the present invention; Figure 3 Flowchart of scene state-driven trajectory prediction; Figure 4 Flowchart for calculating the reliability of a trajectory segment; Figure 5 Flowchart for trajectory segment rendering level decision-making; Figure 6 Schematic diagram of a unified trajectory data model for 2D and 3D maps. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0019] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.

[0020] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0021] like Figure 1 and Figures 3 to 6 As shown, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method according to an embodiment of the present invention includes: Acquire at least two types of target-related spatiotemporal data; The acquired target-related spatiotemporal data are subjected to coordinate system one, time reference unification, and data reliability calibration. A correlation score is calculated based on spatial proximity, temporal proximity, target type consistency, and data source credibility. Data from different data sources are then merged into data for the same target based on the correlation score. A unified trajectory object is constructed based on the fused target data. The unified trajectory object includes at least the target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The importance of a target is determined based on its attributes, abnormal states, trajectory reliability, regional importance, and user interaction behavior. The aggregation score is calculated based on spatial distance, temporal similarity, direction of motion, target importance, and abnormal state, and the multi-target aggregation strategy and layered rendering strategy are determined based on the aggregation score. The rendering level of a trajectory segment is determined based on trajectory reliability, map status, and target importance. The same unified trajectory object is mapped to a two-dimensional map layer and a three-dimensional map entity respectively, and a unified two-dimensional and three-dimensional rendering is performed. The current focus target is determined based on the target's importance, its relevance to the current field of view, alarm status, and user lock status, and the map field of view parameters are automatically adjusted according to the current display mode. When the map is panned, zoomed, rotated, or switched between 2D and 3D, the trajectory loading status, aggregation status, and focus status are updated based on the unified trajectory object.

[0022] According to an embodiment of the present invention, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocus method, in a specific implementation, includes at least two of the following target-related spatiotemporal data: positioning device data, satellite positioning data, mobile terminal location data, target location data output by a video recognition system, business event data, alarm data, map area data, and user interaction behavior data; the acquired individual data are organized into a unified access record R: R = {source_id, target_id, time, lon, lat, x, y, z, speed, heading, type, status, alarm, confidence}, Wherein, source_id represents the data source identifier, target_id represents the target identifier, time represents the data timestamp, lon and lat represent latitude and longitude, x, y, and z represent spatial coordinates, speed represents speed, heading represents direction, type represents target type, status represents behavior status, alarm represents alarm status, and confidence represents data source confidence.

[0023] According to an embodiment of the present invention, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method, in a specific implementation, includes performing coordinate system one by transforming latitude and longitude coordinates, projected coordinates, and three-dimensional coordinates from different sources to a unified spatial coordinate system. The two-dimensional coordinate transformation relationship is as follows: P_2d = T_geo(lon, lat, epsg_source, epsg_target), The three-dimensional coordinate transformation relationship is as follows: P_3d = {x, y, h + h_offset}, Where h_offset is a three-dimensional height correction parameter, which is configured based on map elevation benchmark, device height benchmark, or business scenario. The execution of time benchmark unification includes: time alignment of data from different sources according to a preset time granularity delta_t. t' = floor(t / delta_t) × delta_t, In the real-time display scenario, delta_t is configured to be between 1 and 5 seconds, while in the historical playback scenario, delta_t is configured to be between 5 and 60 seconds.

[0024] In a specific embodiment of the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method according to an embodiment of the present invention, the association score M is calculated in the following manner: M = a1×S_space + a2×S_time + a3×S_type + a4×S_source, a1 + a2 + a3 + a4 = 1, Where S_space represents spatial proximity, S_time represents temporal proximity, S_type represents target type consistency, S_source represents data source credibility, and a1, a2, a3 and a4 are the corresponding weights. When M is greater than the association threshold theta_match, the corresponding data will be merged into the same target. When M is less than or equal to theta_match, a new target candidate object will be generated or it will enter the manual confirmation queue. The associated threshold theta_match ranges from 0 to 1, with a1=0.40, a2=0.25, a3=0.20, and a4=0.15.

[0025] According to an embodiment of the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method of the present invention, in a specific implementation, the unified trajectory object T is represented as: T = {target_id, time_series, point_2d_series, point_3d_series, track_type, confidence_series, importance_score, render_level, focus_state}, Among them, time_series represents time series, point_2d_series represents two-dimensional trajectory point series, point_3d_series represents three-dimensional trajectory point series, track_type represents historical trajectory, real-time trajectory, predicted trajectory or alarm trajectory, confidence_series represents trajectory confidence series, importance_score represents target importance, render_level represents rendering level, and focus_state represents focus state. The trajectory is divided into multiple trajectory segments according to a fixed time window, fixed spatial length, direction change point, or scene boundary. The confidence level C(i,k) is calculated for each trajectory segment. C(i,k) = λ1×S_source(i,k) + λ2×S_continuity(i,k) + λ3×S_motion(i,k) + λ4×S_constraint(i,k), λ1+λ2+λ3+λ4=1, Where S_source represents the reliability of the data source, S_continuity represents the continuity of the trajectory, S_motion represents the kinematic consistency, S_constraint represents the satisfaction of the scene constraints, and λ1=0.30, λ2=0.25, λ3=0.25, λ4=0.20; The reliability of the data source is determined based on the data source level, sampling completeness, and time delay; the trajectory continuity is determined based on the time interval and spatial interval between adjacent trajectory segments; the kinematic consistency is determined based on changes in speed, direction, and altitude; and the scene constraint satisfaction is determined based on constraints such as roads, area boundaries, restricted areas, or business areas.

[0026] According to an embodiment of the present invention, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method, in a specific implementation, the target importance I is calculated in the following manner: I = b1×S_type + b2×S_alarm + b3×S_confidence + b4×S_user + b5×S_region, b1+b2+b3+b4+b5=1, Where S_type represents the target type score, S_alarm represents the anomaly or alarm score, S_confidence represents the trajectory confidence score, S_user represents the user attention behavior score, S_region represents the importance score of the target's region, and b1=0.20, b2=0.30, b3=0.20, b4=0.15, b5=0.15; When I is greater than the first threshold theta_focus, the target is added to the priority target queue; when I is greater than the second threshold theta_alarm, the target is added to the forced focus queue; theta_focus ranges from 0 to 1, theta_alarm ranges from 0 to 1, and theta_alarm is greater than theta_focus.

[0027] According to an embodiment of the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method of the present invention, in a specific implementation, the aggregation score G is calculated in the following manner: G = c1×D_space + c2×D_time + c3×D_heading + c4×(1 - I_norm) + c5×S_normal, c1+c2+c3+c4+c5=1, Wherein, D_space represents the normalized value of spatial distance, D_time represents the normalized value of time difference, D_heading represents the normalized value of orientation difference, I_norm represents the normalized value of target importance, S_normal represents the score of non-abnormal state, and c1=0.30, c2=0.20, c3=0.15, c4=0.25, c5=0.10; Targets with higher importance than theta_focus, abnormal targets, and user-locked targets are not included in normal aggregation; The layered rendering strategy outputs layered rendering results according to the following layers: base layer, common target aggregation layer, common target trajectory layer, key target trajectory layer, abnormal target alarm layer, and user's current focus target layer.

[0028] According to an embodiment of the present invention, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocusing method, in a specific implementation, determines the rendering level L(i,k) as follows: L(i,k) = H(C(i,k), Z, V(i,k), B, I), Where C(i,k) represents the reliability of the trajectory segment, Z represents the map zoom level, V(i,k) represents the positional relationship between the trajectory segment and the current field of view, B represents the current rendering resource status, and I represents the importance of the target; The rendering levels include at least: a first rendering level, which retains the trajectory index and confidence summary but does not generate complete graphic data; a second rendering level, which loads trajectory points at a sampling ratio of 10% to 50% and generates graphic data with an update cycle of 1 to 5 seconds; and a third rendering level, which loads complete trajectory points and generates graphic data with an update cycle of 0.1 to 1 second.

[0029] According to an embodiment of the present invention, the multi-source spatiotemporal data fusion two-dimensional and three-dimensional rendering and target autofocus method, in a specific implementation, the mapping relationship of the unified two-dimensional and three-dimensional rendering is as follows: Render(T, mode) = {Render2D(T.point_2d_series, T.render_level),Render3D(T.point_3d_series, T.render_level)}, Here, mode represents the two-dimensional display mode, the three-dimensional display mode, or the two-dimensional and three-dimensional linkage display mode; two-dimensional rendering uses point_2d_series to generate trajectory lines, trajectory points, target icons, aggregate icons, and alarm markers; three-dimensional rendering uses point_3d_series to generate three-dimensional trajectory lines, three-dimensional target entities, height lines, dynamic trails, and timeline playback objects. The target selection score F is calculated as follows: F = d1×I + d2×S_visible + d3×S_alarm + d4×S_user_lock, When the forced focus queue is not empty, the target with the highest score is selected as the current focus target; when the forced focus queue is empty, the target with the highest overall score is selected as the candidate focus target from the priority target queue; when the user manually locks a target, the user-locked target remains the current focus target until the user unlocks the target or the target becomes invalid. The map view parameters include the map center point, 2D map zoom level, 3D view height, 3D pitch angle, target display level, label display status, and trajectory playback time window.

[0030] Secondly, such as Figure 2 As shown, the multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocus system according to an embodiment of the present invention includes: The data access module is used to acquire at least two types of target-related spatiotemporal data; The spatiotemporal unification module is used to perform coordinate system unification, time reference unification, and data reliability calibration on the acquired target-related spatiotemporal data; The target association and fusion module is used to calculate the association score based on spatial proximity, temporal proximity, target type consistency and data source credibility, and to fuse data from different data sources into data of the same target based on the association score; A unified trajectory object construction module is used to construct a unified trajectory object based on the fused target data. The unified trajectory object includes at least target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The target importance assessment module is used to determine the importance of a target based on its attributes, abnormal status, trajectory reliability, regional importance, and user interaction behavior. The aggregation and rendering strategy decision module is used to calculate the aggregation score based on spatial distance, temporal similarity, motion direction, target importance and abnormal state, and determine the multi-target aggregation strategy and layered rendering strategy based on the aggregation score; The rendering level decision module is used to determine the rendering level of a trajectory segment based on trajectory credibility, map status, and target importance. The unified 2D / 3D rendering module is used to map the same unified trajectory object to a 2D map layer and a 3D map entity respectively, and perform unified 2D / 3D rendering. The autofocus module is used to determine the current focus target based on the target's importance, the correlation between the target and the current field of view, alarm status, and user lock status, and automatically adjust the map field of view parameters according to the current display mode. The status linkage update module is used to update the trajectory loading status, aggregation status, and focus status based on the unified trajectory object when the map is translated, zoomed, rotated, or switched between 2D and 3D.

[0031] To facilitate understanding of the above technical solutions of the present invention, the following detailed description of the above technical solutions of the present invention is provided through specific embodiments and working principles.

[0032] This technical solution is implemented in the following order: multi-source data access, unified spatiotemporal processing, target association and fusion, unified trajectory object construction, trajectory credibility calculation, target importance assessment, aggregation and rendering level decision, unified 2D / 3D rendering, automatic focusing of key targets, and status linkage update.

[0033] Step 1: Acquire spatiotemporal data related to multiple target sources; The system acquires at least two types of target-related spatiotemporal data, including at least two of the following: positioning device data, satellite positioning data, mobile terminal location data, target location data output by the video recognition system, service event data, alarm data, map area data, and user interaction behavior data. The system then organizes each piece of access data into a unified access record R. R = {source_id, target_id, time, lon, lat, x, y, z, speed, heading, type, status, alarm, confidence}, Among them, source_id represents the data source identifier, target_id represents the target identifier, time represents the data timestamp, lon and lat represent latitude and longitude, x, y and z represent spatial coordinates, speed represents speed, heading represents direction, type represents target type, status represents behavior status, alarm represents alarm status, and confidence represents data source confidence.

[0034] Step 2: Perform coordinate system unification, time benchmark unification, and data reliability calibration; The system transforms latitude and longitude coordinates, projected coordinates, and 3D coordinates from different sources to a unified spatial coordinate system. The 2D coordinate transformation relationship is as follows: P_2d = T_geo(lon, lat, epsg_source, epsg_target), The three-dimensional coordinate transformation relationship is as follows: P_3d = {x, y, h + h_offset}, The system performs time alignment on data from different sources according to a preset time granularity delta_t: t' = floor(t / delta_t) × delta_t, The system calibrates the reliability of the data source based on the data source type, sampling completeness, time delay, and spatial offset, resulting in S_source. Here, delta_t is the time alignment granularity, configured from 1 to 5 seconds for real-time display scenarios and from 5 to 60 seconds for historical playback scenarios; h_offset is the 3D height correction parameter, configured based on map elevation benchmarks, device height benchmarks, or business scenario settings.

[0035] Step 3: Perform target association and fusion; The system calculates the association score M based on spatial proximity, temporal proximity, target type consistency, and data source credibility. M = a1×S_space + a2×S_time + a3×S_type + a4×S_source, a1 + a2 + a3 + a4 = 1, When the association score M is greater than the association threshold theta_match, the system merges the corresponding data into the same target; when M is less than or equal to theta_match, the system generates a new target candidate or enters the manual confirmation queue. Theta_match value ranges from 0 to 1, preferably 0.75; the sum of the spatial proximity weight a1, temporal proximity weight a2, target type consistency weight a3, and data source credibility weight a4 is 1, preferably configured as a1=0.40, a2=0.25, a3=0.20, and a4=0.15. Through this step, the location, time, status, and alarm information of the same target in different data sources are merged into unified target data.

[0036] Step 4: Construct a unified trajectory object; The system constructs a unified trajectory object T based on the fused target data: T = {target_id, time_series, point_2d_series, point_3d_series, track_type, confidence_series, importance_score, render_level, focus_state}, Among them, time_series represents time series, point_2d_series represents two-dimensional trajectory point series, point_3d_series represents three-dimensional trajectory point series, track_type represents historical trajectory, real-time trajectory, predicted trajectory or alarm trajectory, confidence_series represents trajectory confidence series, importance_score represents target importance, render_level represents rendering level, and focus_state represents focus state.

[0037] Step 5: Calculate the trajectory confidence sequence; The system divides the trajectory into multiple trajectory segments according to a fixed time window, fixed spatial length, direction change point, or scene boundary, and calculates the confidence level C(i,k) for each trajectory segment: C(i,k) = lambda1×S_source(i,k) + lambda2×S_continuity(i,k) +lambda3×S_motion(i,k) + lambda4×S_constraint(i,k), lambda1 + lambda2 + lambda3 + lambda4 = 1, Wherein, S_source represents data source reliability, S_continuity represents trajectory continuity, S_motion represents kinematic consistency, and S_constraint represents scene constraint satisfaction; lambda1, lambda2, lambda3, and lambda4 are confidence factor weights, with the preferred configuration being lambda1=0.30, lambda2=0.25, lambda3=0.25, and lambda4=0.20. Data source reliability is determined based on data source level, sampling completeness, and time delay; trajectory continuity is determined based on the time interval and spatial interval between adjacent trajectory segments; kinematic consistency is determined based on changes in speed, direction, and altitude; and scene constraint satisfaction is determined based on constraints related to roads, area boundaries, restricted areas, or business areas. The trajectory confidence sequence is written into a unified trajectory object for subsequent target importance assessment, rendering level decision-making, and selection of key targets.

[0038] Step 6: Calculate the importance of the target; The system determines the target importance I based on target attributes, abnormal states, trajectory reliability, business rules, regional importance, and user interaction behavior. I = b1×S_type + b2×S_alarm + b3×S_confidence + b4×S_user + b5×S_region, b1 + b2 + b3 + b4 + b5 = 1, Wherein, S_type represents the target type score, S_alarm represents the anomaly or alarm score, S_confidence represents the trajectory confidence score, S_user represents the user attention behavior score, and S_region represents the importance score of the target's region; the sum of b1, b2, b3, b4, and b5 is 1, with the preferred configuration being b1=0.20, b2=0.30, b3=0.20, b4=0.15, and b5=0.15. When I is greater than the first threshold theta_focus, the system adds the target to the priority target queue; when I is greater than the second threshold theta_alarm, the system adds the target to the forced focus queue. Theta_focus value ranges from 0 to 1, preferably 0.60; theta_alarm value ranges from 0 to 1, preferably 0.85, and theta_alarm is greater than theta_focus.

[0039] Step 7: Determine the multi-target aggregation strategy and the layered rendering strategy; When multiple targets exist within the current field of view, the system calculates an aggregate score G based on spatial distance, temporal similarity, direction of movement, target importance, and abnormal states. G = c1×D_space + c2×D_time + c3×D_heading + c4×(1 - I_norm) + c5×S_normal, Where D_space represents the spatial distance normalization value, D_time represents the time difference normalization value, D_heading represents the direction difference normalization value, I_norm represents the target importance normalization value, and S_normal represents the non-abnormal state score; the sum of c1, c2, c3, c4, and c5 is 1, with the preferred configuration being c1=0.30, c2=0.20, c3=0.15, c4=0.25, and c5=0.10. Targets with a importance higher than theta_focus, abnormal targets, and user-locked targets do not participate in normal aggregation. The system outputs layered rendering results according to the base layer, normal target aggregation layer, normal target trajectory layer, key target trajectory layer, abnormal target alarm layer, and user-current focused target layer.

[0040] Step 8: Determine the rendering level based on trajectory reliability, map status, and target importance; The system obtains the current map view range, map zoom level, 2D or 3D display mode, and rendering resource status, and calculates the trajectory segment rendering level L(i,k): L(i,k) = H(C(i,k), Z, V(i,k), B, I), Where Z represents the map zoom level, V(i,k) represents the positional relationship between the trajectory segment and the current field of view, B represents the current rendering resource status, and I represents the target importance. The rendering levels include at least: First rendering level, retaining the trajectory index and confidence summary, but not generating complete graphic data; Second rendering level, loading trajectory points at a sampling rate of 10% to 50%, and generating graphic data with an update cycle of 1 to 5 seconds; Third rendering level, loading complete trajectory points, and generating graphic data with an update cycle of 0.1 to 1 second. The system can adjust the above sampling rate and update cycle based on device performance, the number of targets within the field of view, and business priority.

[0041] Step 9: Perform a unified 2D / 3D rendering map; The system maps the same uniform trajectory object to both a 2D map layer and a 3D map entity: Render(T, mode) = {Render2D(T.point_2d_series, T.render_level),Render3D(T.point_3d_series, T.render_level)}, Here, `mode` represents the 2D display mode, 3D display mode, or 2D / 3D simultaneous display mode. 2D rendering uses `point_2d_series` to generate trajectory lines, trajectory points, target icons, aggregation icons, and alarm markers; 3D rendering uses `point_3d_series` to generate 3D trajectory lines, 3D target entities, height lines, dynamic wakes, and timeline playback objects.

[0042] Step 10: Execute automatic focusing and status updates for key targets; The system calculates the focus target selection score F based on target importance, the relevance of the target to the current field of view, alarm status, and user lock status. F = d1×I + d2×S_visible + d3×S_alarm + d4×S_user_lock, When the forced focus queue is not empty, the system selects the target with the highest score as the current focus target. When the forced focus queue is empty, the system selects the target with the highest overall score from the priority target queue as the candidate focus target. When the user manually locks a target, the system keeps the user-locked target as the current focus target until the user unlocks it or the target becomes invalid. The system automatically adjusts the map center point, 2D map zoom level, 3D view height, 3D pitch angle, target display level, label display status, and trajectory playback time window according to the current display mode. When the map is panned, zoomed, rotated, or switched between 2D and 3D, the system does not reconstruct the target trajectory data, but updates the trajectory loading status, aggregation status, and focus status based on a unified trajectory object.

[0043] In summary, by utilizing the technical solutions described above, the present invention integrates target data from different sources into a unified data foundation through a unified coordinate system, a unified time reference, and data reliability calibration, reducing the problems of multiple trajectories and inconsistent states for the same target. By encapsulating two-dimensional and three-dimensional trajectory point sequences and rendering states through a unified trajectory object, two-dimensional and three-dimensional maps can be displayed and updated in conjunction with the same data structure. By using trajectory reliability and target importance to participate in rendering level decisions, data loading and graphics updates can be dynamically adjusted according to the target and map states. By excluding key and abnormal targets from ordinary aggregation and setting a layered rendering order, the occlusion of key information by ordinary targets is reduced. By calculating focus scores based on target importance and alarm states and automatically adjusting map view parameters, the workload of manual retrieval and manual view adjustment is reduced. By updating each state based on the unified trajectory object during map switching and view changes, the continuity and consistency of the display results are ensured.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for multi-source spatiotemporal data fusion 2D / 3D rendering and automatic target focusing, characterized in that, include: Acquire at least two types of target-related spatiotemporal data; The acquired target-related spatiotemporal data are subjected to coordinate system one, time reference unification, and data reliability calibration. A correlation score is calculated based on spatial proximity, temporal proximity, target type consistency, and data source credibility. Data from different data sources are then merged into data for the same target based on the correlation score. A unified trajectory object is constructed based on the fused target data. The unified trajectory object includes at least the target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The importance of a target is determined based on its attributes, abnormal states, trajectory reliability, regional importance, and user interaction behavior. The aggregation score is calculated based on spatial distance, temporal similarity, direction of motion, target importance, and abnormal state, and the multi-target aggregation strategy and layered rendering strategy are determined based on the aggregation score. The rendering level of a trajectory segment is determined based on trajectory reliability, map status, and target importance. The same unified trajectory object is mapped to a two-dimensional map layer and a three-dimensional map entity respectively, and a unified two-dimensional and three-dimensional rendering is performed. The current focus target is determined based on the target's importance, its relevance to the current field of view, alarm status, and user lock status, and the map field of view parameters are automatically adjusted according to the current display mode. When the map is panned, zoomed, rotated, or switched between 2D and 3D, the trajectory loading status, aggregation status, and focus status are updated based on the unified trajectory object.

2. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The target-related spatiotemporal data includes at least two of the following: positioning device data, satellite positioning data, mobile terminal location data, target location data output by the video recognition system, service event data, alarm data, map area data, and user interaction behavior data; the acquired individual data entries are organized into a unified access record R: R = {source_id, target_id, time, lon, lat, x, y, z, speed, heading, type, status, alarm, confidence}, Wherein, source_id represents the data source identifier, target_id represents the target identifier, time represents the data timestamp, lon and lat represent latitude and longitude, x, y, and z represent spatial coordinates, speed represents speed, heading represents direction, type represents target type, status represents behavior status, alarm represents alarm status, and confidence represents data source confidence.

3. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, Implementing coordinate system one includes: transforming latitude and longitude coordinates, projected coordinates, and three-dimensional coordinates from different sources to a unified spatial coordinate system. The two-dimensional coordinate transformation relationship is as follows: P_2d = T_geo(lon, lat, epsg_source, epsg_target), The three-dimensional coordinate transformation relationship is as follows: P_3d = {x, y, h + h_offset}, Where h_offset is a three-dimensional height correction parameter, which is configured based on map elevation benchmark, device height benchmark, or business scenario. The execution of time benchmark unification includes: time alignment of data from different sources according to a preset time granularity delta_t. t' = floor(t / delta_t) × delta_t, In the real-time display scenario, delta_t is configured to be between 1 and 5 seconds, while in the historical playback scenario, delta_t is configured to be between 5 and 60 seconds.

4. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The correlation score M is calculated in the following manner: M = a1×S_space + a2×S_time + a3×S_type + a4×S_source, a1 + a2 + a3 + a4 = 1, Where S_space represents spatial proximity, S_time represents temporal proximity, S_type represents target type consistency, S_source represents data source credibility, and a1, a2, a3 and a4 are the corresponding weights. When M is greater than the association threshold theta_match, the corresponding data will be merged into the same target. When M is less than or equal to theta_match, a new target candidate object will be generated or it will enter the manual confirmation queue. The associated threshold theta_match ranges from 0 to 1, with a1=0.40, a2=0.25, a3=0.20, and a4=0.

15.

5. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The unified trajectory object T is represented as: T = {target_id, time_series, point_2d_series, point_3d_series, track_type, confidence_series, importance_score, render_level, focus_state}, Among them, time_series represents time series, point_2d_series represents two-dimensional trajectory point series, point_3d_series represents three-dimensional trajectory point series, track_type represents historical trajectory, real-time trajectory, predicted trajectory or alarm trajectory, confidence_series represents trajectory confidence series, importance_score represents target importance, render_level represents rendering level, and focus_state represents focus state. The trajectory is divided into multiple trajectory segments according to a fixed time window, fixed spatial length, direction change point, or scene boundary. The confidence level C(i,k) is calculated for each trajectory segment. C(i,k) = λ1×S_source(i,k) + λ2×S_continuity(i,k) + λ3×S_motion(i,k) +λ4×S_constraint(i,k), λ1+λ2+λ3+λ4=1, Where S_source represents the reliability of the data source, S_continuity represents the continuity of the trajectory, S_motion represents the kinematic consistency, S_constraint represents the satisfaction of the scene constraints, and λ1=0.30, λ2=0.25, λ3=0.25, λ4=0.20; The reliability of the data source is determined based on the data source level, sampling completeness, and time delay; the trajectory continuity is determined based on the time interval and spatial interval between adjacent trajectory segments; the kinematic consistency is determined based on changes in speed, direction, and altitude; and the scene constraint satisfaction is determined based on constraints such as roads, area boundaries, restricted areas, or business areas.

6. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The target importance I is calculated as follows: I = b1×S_type + b2×S_alarm + b3×S_confidence + b4×S_user + b5×S_region, b1+b2+b3+b4+b5=1, Where S_type represents the target type score, S_alarm represents the anomaly or alarm score, S_confidence represents the trajectory confidence score, S_user represents the user attention behavior score, S_region represents the importance score of the target's region, and b1=0.20, b2=0.30, b3=0.20, b4=0.15, b5=0.15; When I is greater than the first threshold theta_focus, the target is added to the priority target queue; when I is greater than the second threshold theta_alarm, the target is added to the forced focus queue; theta_focus ranges from 0 to 1, theta_alarm ranges from 0 to 1, and theta_alarm is greater than theta_focus.

7. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The aggregate score G is calculated as follows: G = c1×D_space + c2×D_time + c3×D_heading + c4×(1 - I_norm) + c5×S_normal, c1+c2+c3+c4+c5=1, Wherein, D_space represents the normalized value of spatial distance, D_time represents the normalized value of time difference, D_heading represents the normalized value of orientation difference, I_norm represents the normalized value of target importance, S_normal represents the score of non-abnormal state, and c1=0.30, c2=0.20, c3=0.15, c4=0.25, c5=0.10; Targets with higher importance than theta_focus, abnormal targets, and user-locked targets are not included in normal aggregation; The layered rendering strategy outputs layered rendering results according to the following layers: base layer, common target aggregation layer, common target trajectory layer, key target trajectory layer, abnormal target alarm layer, and user's current focus target layer.

8. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The rendering level L(i,k) is determined as follows: L(i,k) = H(C(i,k), Z, V(i,k), B, I), Where C(i,k) represents the reliability of the trajectory segment, Z represents the map zoom level, V(i,k) represents the positional relationship between the trajectory segment and the current field of view, B represents the current rendering resource status, and I represents the importance of the target; The rendering levels include at least: a first rendering level, which retains the trajectory index and confidence summary but does not generate complete graphic data; a second rendering level, which loads trajectory points at a sampling ratio of 10% to 50% and generates graphic data with an update cycle of 1 to 5 seconds; and a third rendering level, which loads complete trajectory points and generates graphic data with an update cycle of 0.1 to 1 second.

9. The method for multi-source spatiotemporal data fusion 2D / 3D rendering and target autofocusing according to claim 1, characterized in that, The mapping relationship for the unified 2D / 3D rendering is as follows: Render(T, mode) = {Render2D(T.point_2d_series, T.render_level), Render3D(T.point_3d_series, T.render_level)}, Here, mode represents the two-dimensional display mode, the three-dimensional display mode, or the two-dimensional and three-dimensional linkage display mode; two-dimensional rendering uses point_2d_series to generate trajectory lines, trajectory points, target icons, aggregate icons, and alarm markers; three-dimensional rendering uses point_3d_series to generate three-dimensional trajectory lines, three-dimensional target entities, height lines, dynamic trails, and timeline playback objects. The target selection score F is calculated as follows: F = d1×I + d2×S_visible + d3×S_alarm + d4×S_user_lock, When the forced focus queue is not empty, the target with the highest score is selected as the current focus target; when the forced focus queue is empty, the target with the highest overall score is selected as the candidate focus target from the priority target queue; when the user manually locks a target, the user-locked target remains the current focus target until the user unlocks the target or the target becomes invalid. The map view parameters include the map center point, 2D map zoom level, 3D view height, 3D pitch angle, target display level, label display status, and trajectory playback time window.

10. A multi-source spatiotemporal data fusion system for 2D / 3D rendering and automatic target focusing, characterized in that: include: The data access module is used to acquire at least two types of target-related spatiotemporal data; The spatiotemporal unification module is used to perform coordinate system unification, time reference unification, and data reliability calibration on the acquired target-related spatiotemporal data; The target association and fusion module is used to calculate the association score based on spatial proximity, temporal proximity, target type consistency and data source credibility, and to fuse data from different data sources into data of the same target based on the association score; A unified trajectory object construction module is used to construct a unified trajectory object based on the fused target data. The unified trajectory object includes at least target identifier, time series, two-dimensional trajectory point sequence, three-dimensional trajectory point sequence, trajectory type, trajectory confidence sequence, target importance, rendering level, and focus state. The target importance assessment module is used to determine the importance of a target based on its attributes, abnormal status, trajectory reliability, regional importance, and user interaction behavior. The aggregation and rendering strategy decision module is used to calculate the aggregation score based on spatial distance, temporal similarity, motion direction, target importance and abnormal state, and determine the multi-target aggregation strategy and layered rendering strategy based on the aggregation score; The rendering level decision module is used to determine the rendering level of a trajectory segment based on trajectory credibility, map status, and target importance. The unified 2D / 3D rendering module is used to map the same unified trajectory object to a 2D map layer and a 3D map entity respectively, and perform unified 2D / 3D rendering. The autofocus module is used to determine the current focus target based on the target's importance, the correlation between the target and the current field of view, alarm status, and user lock status, and automatically adjust the map field of view parameters according to the current display mode. The status linkage update module is used to update the trajectory loading status, aggregation status, and focus status based on the unified trajectory object when the map is translated, zoomed, rotated, or switched between 2D and 3D.

Images