CN-121723110-B - Method and system for monitoring natural resource area

CN121723110BCN 121723110 BCN121723110 BCN 121723110BCN-121723110-B

Abstract

The application provides a monitoring method and a system of a natural resource area, which relate to the technical field of natural resource monitoring, wherein a spectrum characteristic image layer is generated by multi-source remote sensing data covering the monitoring area, the ground surface polarization information of a ground monitoring device is synchronously acquired, a polarization characteristic image is generated, and the reflection characteristics of natural objects and artifacts are distinguished; and according to the continuously acquired spectral feature layer and polarization feature map, combining spectral feature transformation and artifact key region expansion, and judging whether the potential illegal map spots are continuously transformed or not. And the related digital orthographic images and the boundaries of the pattern spots are sleeved on the monitoring video picture, so that the accurate positioning of the pattern spots in the natural resource region and the dynamic identification of the target activity can be realized.

Inventors

CHEN XIANG
ZHAI JIAN
Qin Yikun
DONG QUANJIE
YIN LAN
LIANG XIAOYAN
MA HUIQIANG
GUO ZHONGYING
TONG SUO

Assignees

北京威特空间科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260210

Claims (9)

1. A method for monitoring a natural resource region, comprising: Generating a spectrum characteristic layer according to the acquired multi-source remote sensing data covering the monitoring area, wherein the spectrum characteristic layer is used for representing spectrum characteristic differences of natural objects in the monitoring area from a first state to a second state; synchronously acquiring surface polarization information acquired by a ground monitoring device arranged in the monitoring area, and processing to generate a polarization characteristic diagram, wherein the polarization characteristic diagram is used for distinguishing the reflection characteristic of a natural object from the reflection characteristic of an artificial object; Fusing the spectral features of the state change of the natural objects identified from the spectral feature layer with the key areas with the reflection characteristics of artifacts identified from the polarization feature layer, and locating potential illegal patterns in the monitoring areas; Based on the continuously acquired spectral feature layer and the polarization feature map, judging whether the potential violation pattern is continuously transited from a first state to a second state, wherein the judgment of the transition is required to be simultaneously based on the transition of the spectral feature and the expansion of a key area with artifacts; when the continuous transition is judged to occur, generating a target violation pattern and corresponding geospatial information; based on the geographic space information of the target violation pattern, sleeving the associated digital orthographic image and the pattern boundary corresponding to the target violation pattern into a monitoring video picture at a corresponding position, and dynamically identifying target activity in the pattern boundary; Wherein after locating the potential violation pattern within the surveillance area, the step of determining whether a sustained transition of the potential violation pattern from a first state to a second state occurs further comprises: Acquiring night light data covering the potential violation pattern area; analyzing the night light data to identify whether an artificial light abnormal region exists in the boundary of the potential violation pattern, wherein the artificial light abnormal region exists in the boundary, and the brightness value and the distribution range are stable or increase along with time; Taking the abnormal area with the artificial lamplight as an additional verification condition for judging that the continuous transition occurs; When one potentially offending pattern meets both the continuous transition of the spectral feature, the enlargement of the critical area with artifacts, and the presence of the artificial light anomaly, it is determined that a continuous transition from the first state to the second state occurs.
2. The method of claim 1, wherein said fusing spectral features of state changes of artifacts identified from said spectral feature map layer with key regions of reflective properties of artifacts identified from said polarization feature map locates potential violation map plaques within said monitored regions, comprising: comparing spectral characteristic layers of different time phases by adopting a change vector analysis algorithm, and identifying pixels with obviously reduced vegetation indexes as primary candidate areas of state change by calculating vegetation indexes representing vegetation coverage change amounts; Meanwhile, linear polarization degree is calculated on the earth surface polarization information, pixels with high linear polarization degree are screened out through setting a threshold value, and a key area with reflection characteristics of artifacts is generated; And carrying out spatial superposition analysis on the preliminary candidate region and the key region through a weighted decision fusion model, and clustering overlapping regions which simultaneously meet two conditions of vegetation index significant reduction and high linear polarization degree so as to locate potential violation pattern spots in the monitoring region.
3. The method of claim 2, wherein the determining whether the persistent transition of the potentially offending plaque from the first state to the second state occurs based on the continuously acquired spectral signature and the polarization signature comprises: Establishing a spectrum and polarization characteristic file of each potential violation pattern spot on a continuous time sequence, and judging whether the vegetation coverage of the potential violation pattern spot shows a descending trend or not by carrying out trend analysis on the vegetation index mean value in the pattern spot; calculating the area of a key area with the reflection characteristics of artifacts in the potential violation pattern spots, and judging whether the area is enlarged in time sequence or not; the transition process is confirmed to be a persistent transition if and only if one potentially offending pattern meets both a continuous drop in vegetation coverage and a significant expansion in the area of the critical area of the artifact.
4. The method of claim 1, wherein the dynamically identifying target activity within the plaque boundary comprises: Extracting pixel areas positioned in the image spot boundaries from the continuous monitoring video pictures; Comparing pixel value differences of the pixel areas in adjacent video frames to obtain difference areas representing pixel value changes; Determining at least one independently moving target according to the association relation between the position and the shape of the difference area in the picture in continuous multiframes, and tracking the continuous position of the moving target in the picture; Extracting the shape, texture and motion pattern characteristics of the moving object presented in a video picture, and classifying the moving object into a predefined dynamic object type based on a preset characteristic comparison rule; Analyzing the change rule of the tracked continuous position of the moving target along with time, and judging the behavior type of the moving target according to a preset behavior mode rule; and when at least one of the dynamic target type and the behavior type does not accord with a preset violation judgment condition, generating a task to be manually checked.
5. The method of claim 1, wherein generating the spectral signature layer comprises: extracting reflection information of red light and near infrared bands which are sensitive to vegetation states from the first spectrum image and the second spectrum image of the multi-source remote sensing data; calculating a vegetation index based on the reflection information, wherein the vegetation index is used for quantifying the surface vegetation coverage condition; And extracting a region with obviously reduced vegetation index in the monitoring region through change vector analysis based on the change of the vegetation index in time sequence, so as to form a spectrum characteristic layer.
6. The method of claim 1, wherein the synchronizing acquires surface polarization information acquired by surface monitoring devices deployed in the monitoring area and processing to generate a polarization signature comprises: Analyzing Stokes vectors from the surface polarization information, and calculating to obtain linear polarization degree; Setting a threshold value for the linear polarization degree, and identifying pixels higher than the threshold value as high-directivity reflection areas, wherein the high-directivity reflection areas correspond to the key areas with the reflection characteristics of artifacts; And generating a polarization characteristic diagram based on the high-directivity reflection area.
7. The method of claim 1, wherein generating the target violation pattern and corresponding geospatial information when the persistent transition is determined to occur comprises: Marking potential violation pattern spots meeting two conditions of continuous reduction of vegetation coverage and significant expansion of artifacts about area of the area as target violation pattern spots; Extracting a contour boundary coordinate set of the target violation pattern spot under a geographic coordinate system of a monitoring area; and associating the outline boundary coordinate set with the unique identifier of the target violation pattern spot to generate geospatial information.
8. The method according to claim 1, wherein the registering the associated digital orthographic image and the map spot boundary corresponding to the target violation map spot to the surveillance video frame of the corresponding location based on the geospatial information of the target violation map spot comprises: According to the geospatial information, invoking a video stream of a monitoring camera, wherein the observation range of the monitoring camera covers the target violation pattern; and overlapping the digital orthographic image corresponding to the target violation pattern spot and the pattern spot boundary into a real-time image frame of the video stream, wherein the digital orthographic image is obtained by shooting a monitoring area through a scheduling unmanned aerial vehicle.
9. A system for monitoring a natural resource area, comprising: The acquisition module is used for generating a spectrum characteristic layer according to the acquired multi-source remote sensing data covering the monitoring area, wherein the spectrum characteristic layer is used for representing spectrum characteristic differences of natural objects in the monitoring area from a first state to a second state; The ground surface polarization information collected by the ground monitoring device arranged in the monitoring area is synchronously obtained, and a polarization characteristic diagram is generated by processing, wherein the polarization characteristic diagram is used for distinguishing the reflection characteristic of a natural object from the reflection characteristic of an artificial object; The fusion module is used for fusing the spectral characteristics of the state change of the natural objects identified in the spectral characteristic layer and the key areas with the reflection characteristics of artifacts identified in the polarization characteristic layer, and locating potential illegal patterns in the monitoring areas; The judging module is used for judging whether the potential violation pattern is continuously converted from the first state to the second state or not based on the continuously acquired spectral feature layer and the polarization feature map, wherein the conversion is judged according to the conversion of the spectral features and the expansion of a key area with artifacts at the same time; The generation module is used for generating target violation pattern spots and corresponding geographic space information when the continuous transition is judged to occur; And the identification module is used for sleeving the associated digital orthographic image and the map spot boundary corresponding to the target violation map spot into a monitoring video picture at a corresponding position based on the geographic space information of the target violation map spot, and dynamically identifying the target activity in the map spot boundary.

Description

Method and system for monitoring natural resource area Technical Field The present application relates to the field of natural resource monitoring technologies, and in particular, to a method and a system for monitoring a natural resource area. Background Natural resource area monitoring is a core technical means of national soil management and control and ecological protection, and abnormal resource utilization conditions can be found in time. The technology is widely applied to the fields of farmland protection, ecological red-line control and the like, and has important significance for maintaining sustainable utilization of natural resources. At present, single optical remote sensing technology is mostly adopted in the existing natural resource area monitoring, and suspected illegal patterns are identified by acquiring multi-temporal remote sensing images and comparing spectral changes covered by analysis ground surfaces. The method has the advantages of wide coverage range and convenient data acquisition, and becomes one of the current mainstream monitoring means. However, the single optical remote sensing technology is difficult to effectively distinguish the state change of a natural object from the change caused by intervention of an artificial object (such as illegal operation of an excavator, construction of an illegal building and the like), and natural evolution is easily misjudged as illegal behaviors, or illegal conditions of hidden intervention of the artificial object are omitted, so that the monitoring result is misdetected, and the omission rate is higher. Therefore, the technical problem of insufficient accuracy of monitoring the natural resource area violations exists in the prior art. Disclosure of Invention The application aims to provide a natural resource area monitoring method and system, which are used for solving the problem of insufficient accuracy of monitoring the natural resource area violations in the prior art. In order to solve the above technical problems, in a first aspect, the present application provides a method for monitoring a natural resource area, including: Generating a spectrum characteristic layer according to the acquired multi-source remote sensing data covering the monitoring area, wherein the spectrum characteristic layer is used for representing spectrum characteristic differences of natural objects in the monitoring area from a first state to a second state; synchronously acquiring surface polarization information acquired by a ground monitoring device arranged in the monitoring area, and processing to generate a polarization characteristic diagram, wherein the polarization characteristic diagram is used for distinguishing the reflection characteristic of a natural object from the reflection characteristic of an artificial object; Fusing the spectral features of the state change of the natural objects identified from the spectral feature layer with the key areas with the reflection characteristics of artifacts identified from the polarization feature layer, and locating potential illegal patterns in the monitoring areas; Based on the continuously acquired spectral feature layer and the polarization feature map, judging whether the potential violation pattern is continuously transited from a first state to a second state, wherein the judgment of the transition is required to be simultaneously based on the transition of the spectral feature and the expansion of a key area with artifacts; when the continuous transition is judged to occur, generating a target violation pattern and corresponding geospatial information; Based on the geographic space information of the target violation pattern, the related digital orthographic image and the pattern boundary corresponding to the target violation pattern are sleeved into a monitoring video picture at a corresponding position, and target activities are dynamically identified in the pattern boundary. Optionally, after locating the potentially offending pattern within the monitored area, the step of determining whether a sustained transition of the potentially offending pattern from the first state to the second state occurs further comprises: Acquiring night light data covering the potential violation pattern area; analyzing the night light data to identify whether an artificial light abnormal region exists in the boundary of the potential violation pattern, wherein the artificial light abnormal region exists in the boundary, and the brightness value and the distribution range are stable or increase along with time; Taking the abnormal area with the artificial lamplight as an additional verification condition for judging that the continuous transition occurs; When one potentially offending pattern meets both the continuous transition of the spectral feature, the enlargement of the critical area with artifacts, and the presence of the artificial light anomaly, it is determined that a continuous transition from the fir