CN-121999432-A - Target object searching method, device, electronic equipment and medium

CN121999432ACN 121999432 ACN121999432 ACN 121999432ACN-121999432-A

Abstract

The embodiment of the application discloses a target object searching method, a target object searching device, electronic equipment and a storage medium, which comprise the steps of obtaining characteristic data of a target object to be searched and a reference monitoring camera corresponding to the target object to be searched for the first time, determining a candidate monitoring camera set based on a monitoring range of the reference monitoring camera, track characteristics of the target to be searched and a space-time distribution relation of each monitoring camera, executing hierarchical characteristic matching on image data collected by the candidate monitoring camera set, and generating a searching result corresponding to the target object to be searched according to a matching result and a coverage area of the candidate monitoring camera set, so that the efficiency and the precision of target object searching can be improved, and the searching effect is improved.

Inventors

CHEN LUBIN
YAN YONGLONG
CHEN LIANG

Assignees

深圳市小鹰视界智能有限公司

Dates

Publication Date: 20260508
Application Date: 20260106

Claims (10)

1. A target object searching method, comprising: acquiring characteristic data of a target object to be searched and capturing a reference monitoring camera corresponding to the target object to be searched for the first time; Determining a candidate monitoring camera set based on the monitoring range of the reference monitoring camera, the track characteristics of the target to be searched and the space-time distribution relation of each monitoring camera; performing hierarchical feature matching on the image data collected by the candidate monitoring camera set; and generating a search result corresponding to the target object to be searched according to the matching result and the coverage range of the candidate monitoring camera set.
2. The target object searching method according to claim 1, wherein the determining the candidate monitoring camera group based on the monitoring range of the reference monitoring camera, the track characteristics of the target to be searched, and the spatio-temporal distribution relation of each monitoring camera includes: extracting track fragments of a target object to be searched in a monitoring range of a reference monitoring camera; acquiring the installation position, the monitoring angle and the coverage data of each monitoring camera, and calculating the space distance between each camera and the reference monitoring camera and the connection degree of the monitoring range; And determining a candidate monitoring camera set according to the track segments, the space distance between each camera and the reference monitoring camera and the connection degree of the monitoring range.
3. The method according to claim 2, wherein determining the candidate monitoring camera group according to the track segments and the spatial distance and the monitoring range engagement between each camera and the reference monitoring camera comprises: acquiring the track direction, the motion speed and the track curvature of the target object to be searched from the track segment; constructing track characteristics of the target object to be searched according to the track direction, the motion speed and the track curvature; Predicting a predicted motion path of the target object to be searched based on the track characteristics, the space distance between each camera and the reference monitoring camera and the monitoring range connection degree; and determining the monitoring camera with the intersection of the predicted motion path and the monitoring range to obtain a candidate monitoring camera set.
4. The method for searching for a target object according to claim 2, wherein the obtaining the installation position, the monitoring angle and the coverage data of each monitoring camera, and calculating the spatial distance between each camera and the reference monitoring camera and the degree of engagement between each camera and the monitoring range, comprises: Calculating the overlapping length of the boundary between the reference monitoring camera and the camera to be calculated according to the boundary coordinates of the monitoring range of the reference monitoring camera and the boundary coordinates of the monitoring range of the camera to be calculated. Calculating the average circumference corresponding to the monitoring range according to the circumference of the monitoring range of the reference monitoring camera and the circumference of the monitoring range of the camera to be calculated; And calculating the ratio of the overlapping length to the average perimeter to obtain the monitoring range connection degree of the camera to be calculated and the reference monitoring camera.
5. The target object searching method of claim 1, wherein said performing hierarchical feature matching on image data collected by said candidate monitoring camera set comprises: extracting first image features of image data acquired by the candidate monitoring camera set; Determining a candidate image set in image data collected by the candidate monitoring camera set according to the object characteristics and the first image characteristics of the target object to be searched; Acquiring a second image feature from the candidate image set; And determining a target image set of the target object to be searched in the candidate image set according to the object features and the second image features.
6. The target object searching method according to claim 5, wherein the determining a candidate image set from the image data collected by the candidate monitoring camera set according to the object feature and the first image feature of the target object to be searched comprises: calculating the similarity between the object characteristics of the target object to be searched and the first image characteristics; and removing images with similarity smaller than the first preset similarity from the image data collected by the candidate monitoring camera group to obtain a candidate image set.
7. The target object searching method according to claim 5, wherein the determining the target image set of the target object to be searched in the candidate image set according to the object feature and the second image feature includes: calculating the similarity between the object characteristics of the target object to be searched and the second image characteristics; And determining images with similarity greater than or equal to a second preset similarity in the candidate image set to obtain a target image set of the target object to be searched.
8. A target object search apparatus, comprising: The acquisition module is used for acquiring characteristic data of a target object to be searched and a reference monitoring camera corresponding to the target object to be searched for the first time; The determining module is used for determining a candidate monitoring camera set based on the monitoring range of the reference monitoring camera, the track characteristics of the target to be searched and the space-time distribution relation of each monitoring camera; the matching module is used for performing hierarchical feature matching on the image data collected by the candidate monitoring camera set; and the generating module is used for generating a search result corresponding to the target object to be searched according to the matching result and the coverage range of the candidate monitoring camera set.
9. An electronic device comprising a memory, a processor and a processor program stored on the memory and executable on the processor, wherein the processor executes the program steps of the target object searching method according to any one of claims 1 to 7.
10. A storage medium storing a computer processing program capable of being loaded by a processor and executing the target object searching method according to any one of claims 1 to 7.

Description

Target object searching method, device, electronic equipment and medium Technical Field The present application relates to the field of monitoring technologies, and in particular, to a target object searching method, a device, an electronic apparatus, and a storage medium. Background In many fields such as security monitoring, traffic control and public area management, tracking and searching of specific target objects (such as suspects, criminal vehicles, lost persons and the like) are often required based on a monitoring system. Along with the continuous perfection of the monitoring network, the deployment range of the monitoring cameras is increasingly wide, the number is continuously increased, and how to quickly and accurately position the target object in the collected data of massive monitoring cameras becomes a key difficult problem in the application of the current monitoring technology. In the prior art, a full-scale camera traversal matching mode is adopted for target object searching, so that the efficiency is low, matching precision is insufficient easily caused by irrelevant data interference, and meanwhile, when a candidate monitoring range is determined, the space-time distribution relation between the motion characteristics of a target object and a camera is often not fully combined, so that the candidate range is overlarge or a key camera is omitted, the searching effect is further influenced, and the problems of low efficiency and poor precision and poor searching effect are solved in the existing target object searching scheme. Disclosure of Invention The embodiment of the application provides a target object searching method, electronic equipment, device and storage medium, which can improve the efficiency and precision of target object searching, thereby improving the searching effect. In a first aspect, an embodiment of the present application provides a target object searching method, including: acquiring characteristic data of a target object to be searched and capturing a reference monitoring camera corresponding to the target object to be searched for the first time; Determining a candidate monitoring camera set based on the monitoring range of the reference monitoring camera, the track characteristics of the target to be searched and the space-time distribution relation of each monitoring camera; performing hierarchical feature matching on the image data collected by the candidate monitoring camera set; and generating a search result corresponding to the target object to be searched according to the matching result and the coverage range of the candidate monitoring camera set. Optionally, in some embodiments of the present application, the determining the candidate monitoring camera set based on the monitoring range of the reference monitoring camera, the track feature of the target to be searched, and the spatial-temporal distribution relation of each monitoring camera includes: extracting track fragments of a target object to be searched in a monitoring range of a reference monitoring camera; acquiring the installation position, the monitoring angle and the coverage data of each monitoring camera, and calculating the space distance between each camera and the reference monitoring camera and the connection degree of the monitoring range; And determining a candidate monitoring camera set according to the track segments, the space distance between each camera and the reference monitoring camera and the connection degree of the monitoring range. Optionally, in some embodiments of the present application, the determining the candidate monitoring camera set according to the track segment, the spatial distance between each camera and the reference monitoring camera, and the connection degree of the monitoring range includes: acquiring the track direction, the motion speed and the track curvature of the target object to be searched from the track segment; constructing track characteristics of the target object to be searched according to the track direction, the motion speed and the track curvature; Predicting a predicted motion path of the target object to be searched based on the track characteristics, the space distance between each camera and the reference monitoring camera and the monitoring range connection degree; and determining the monitoring camera with the intersection of the predicted motion path and the monitoring range to obtain a candidate monitoring camera set. Optionally, in some embodiments of the present application, the acquiring the installation position, the monitoring angle and the coverage data of each monitoring camera, and calculating the spatial distance between each camera and the reference monitoring camera and the connection degree of the monitoring range include: Calculating the overlapping length of the boundary between the reference monitoring camera and the camera to be calculated according to the boundary coordinates of the monitoring range of the ref