CN-121999447-A - Traffic abnormal event detection method, device, storage medium, program product and computer equipment

CN121999447ACN 121999447 ACN121999447 ACN 121999447ACN-121999447-A

Abstract

The application discloses a traffic abnormal event detection method, a device, a storage medium, a program product and computer equipment, wherein the method comprises the steps of obtaining multi-mode data corresponding to a target area, wherein the multi-mode data comprises point cloud data and video data; the method comprises the steps of aligning multi-mode data based on alignment parameters, carrying out target detection on the basis of the aligned multi-mode data to obtain a point cloud detection result and a video detection result, fusing the aligned multi-mode data based on the point cloud detection result and the video detection result to obtain fused data, and carrying out traffic abnormal event detection on the basis of the fused data to obtain a traffic abnormal event detection result, so that the accuracy of the traffic abnormal event detection can be improved.

Inventors

HUANG CHAOPING
CHEN JINXUAN
LIU JINING
SONG KUN
HUANG XINGWEI
WANG YANG
WEN SHANGDONG
WU SHIFAN
XU HAO
LIN JIAO

Assignees

中国移动通信集团广东有限公司
中移湾区(广东)创新研究院有限公司
中国移动通信集团有限公司

Dates

Publication Date: 20260508
Application Date: 20260106

Claims (10)

1. A traffic anomaly detection method, comprising: Acquiring multi-mode data corresponding to a target area, wherein the multi-mode data comprises point cloud data and video data; Aligning the multi-modal data based on an alignment parameter; performing target detection based on the aligned multi-mode data to obtain a point cloud detection result and a video detection result; Based on the point cloud detection result and the video detection result, fusing the aligned multi-mode data to obtain fused data; And detecting the traffic abnormal event based on the fusion data to obtain a traffic abnormal event detection result.
2. The method according to claim 1, wherein the alignment parameters comprise time parameters, calibration parameters and/or reference points corresponding to the target area; The method for acquiring the calibration parameters comprises the following steps: Mapping video data in the multi-mode data corresponding to the target area to a coordinate system where point cloud data in the multi-mode data corresponding to the target area is located to obtain video mapping data; and determining the calibration parameters by using a least square method based on the video mapping data and point cloud data in the multi-mode data corresponding to the target area.
3. The method of claim 1, wherein the fusing the aligned multi-modal data based on the point cloud detection result and the video detection result to obtain fused data comprises: mapping the point cloud detection result to an image coordinate system corresponding to the video detection result to obtain a point cloud detection mapping result; and fusing the aligned multi-mode data based on the point cloud detection mapping result and the video detection result to obtain fused data.
4. The method of claim 3, wherein fusing the aligned multi-modal data based on the point cloud detection mapping result and the video detection result to obtain fused data comprises: Determining an intersection ratio IOU matrix set based on the point cloud detection mapping result and the video detection result; Based on the IOU matrix set, determining a global optimal matching result by using a Hungary algorithm; And fusing the aligned multi-mode data based on the global optimal matching result and a preset IOU threshold value to obtain fused data.
5. The method of claim 1, wherein the detecting the traffic anomaly event based on the fused data to obtain a traffic anomaly event detection result comprises: constructing corresponding characteristic variables based on the fusion data; and calling an abnormal event detector to detect traffic abnormal events based on the characteristic variables, and obtaining a traffic abnormal event detection result.
6. The method according to any one of claims 1-5, further comprising: and obtaining abnormal event evidence obtaining information corresponding to the traffic abnormal event detection result.
7. A traffic anomaly event detection device, comprising: The multi-mode data acquisition module is used for acquiring multi-mode data corresponding to the target area, wherein the multi-mode data comprises point cloud data and video data; the alignment module is used for aligning the multi-mode data based on the alignment parameters; the target detection module is used for carrying out target detection based on the aligned multi-mode data to obtain a point cloud detection result and a video detection result; the fusion module is used for fusing the aligned multi-mode data based on the point cloud detection result and the video detection result to obtain fusion data; And the event detection module is used for detecting traffic abnormal events based on the fusion data to obtain traffic abnormal event detection results.
8. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the method of any of claims 1-6.
9. A computer program product comprising computer instructions which, when executed by a processor, implement the method of any of claims 1-6.
10. A computer device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the method of any of claims 1-6 when the computer program is executed.

Description

Traffic abnormal event detection method, device, storage medium, program product and computer equipment Technical Field The present application relates to the field of data processing technologies, and in particular, to a traffic abnormal event detection method, a device, a storage medium, a program product, and a computer apparatus. Background In the related art, it is common to determine whether a traffic abnormality has occurred by analyzing a video picture taken by a camera. However, the pictures taken by the camera under some special conditions (e.g. severe weather conditions, night, insufficient light) are not clear enough, and are difficult to use for performing accurate event analysis, resulting in low accuracy of traffic anomaly detection. Disclosure of Invention In order to solve the technical problems, the embodiment of the application provides a traffic abnormal event detection method, a device, a storage medium, a program product and computer equipment, which can improve the accuracy of traffic abnormal event detection. In a first aspect, an embodiment of the present application provides a traffic anomaly detection method, including: Acquiring multi-mode data corresponding to a target area, wherein the multi-mode data comprises point cloud data and video data; Aligning the multi-modal data based on an alignment parameter; performing target detection based on the aligned multi-mode data to obtain a point cloud detection result and a video detection result; Based on the point cloud detection result and the video detection result, fusing the aligned multi-mode data to obtain fused data; And detecting the traffic abnormal event based on the fusion data to obtain a traffic abnormal event detection result. Optionally, the alignment parameter includes a time parameter, a calibration parameter and/or a reference point corresponding to the target area; The method for acquiring the calibration parameters comprises the following steps: Mapping video data in the multi-mode data corresponding to the target area to a coordinate system where point cloud data in the multi-mode data corresponding to the target area is located to obtain video mapping data; and determining the calibration parameters by using a least square method based on the video mapping data and point cloud data in the multi-mode data corresponding to the target area. Optionally, the fusing the aligned multi-mode data based on the point cloud detection result and the video detection result to obtain fused data includes: mapping the point cloud detection result to an image coordinate system corresponding to the video detection result to obtain a point cloud detection mapping result; and fusing the aligned multi-mode data based on the point cloud detection mapping result and the video detection result to obtain fused data. Optionally, the fusing the aligned multi-mode data based on the point cloud detection mapping result and the video detection result to obtain fused data includes: Determining an intersection ratio IOU matrix set based on the point cloud detection mapping result and the video detection result; Based on the IOU matrix set, determining a global optimal matching result by using a Hungary algorithm; And fusing the aligned multi-mode data based on the global optimal matching result and a preset IOU threshold value to obtain fused data. Optionally, the detecting the traffic abnormal event based on the fusion data to obtain a traffic abnormal event detection result includes: constructing corresponding characteristic variables based on the fusion data; and calling an abnormal event detector to detect traffic abnormal events based on the characteristic variables, and obtaining a traffic abnormal event detection result. Optionally, the method further comprises: and obtaining abnormal event evidence obtaining information corresponding to the traffic abnormal event detection result. In a second aspect, an embodiment of the present application provides a traffic anomaly event detection apparatus, including: The multi-mode data acquisition module is used for acquiring multi-mode data corresponding to the target area, wherein the multi-mode data comprises point cloud data and video data; the alignment module is used for aligning the multi-mode data based on the alignment parameters; the target detection module is used for carrying out target detection based on the aligned multi-mode data to obtain a point cloud detection result and a video detection result; the fusion module is used for fusing the aligned multi-mode data based on the point cloud detection result and the video detection result to obtain fusion data; And the event detection module is used for detecting traffic abnormal events based on the fusion data to obtain traffic abnormal event detection results. In a third aspect, embodiments of the present application provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed b