Object detection method, device, electronic device and storage medium
By extracting and utilizing multi-dimensional target feature information for object detection, the problem of insufficient accuracy in complex scenarios in the prior art is solved, and higher detection accuracy and positioning accuracy are achieved.
CN114372999BActive Publication Date: 2025-03-21ZHEJIANG DAHUA TECH CO LTD
2 Cites 0 Cited by
Patent Information
- Application Number
- CN202111566878.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-20
- Publication Date
- 2025-03-21
- Estimated Expiration
- 2041-12-20
AI Technical Summary
Technical Problem
The existing single-objective tracking algorithm is difficult to accurately determine the target and background in complex scenarios, and does not fully consider the scale of the target, resulting in insufficient detection accuracy.
Method used
By acquiring the initial video frame of the video to be detected and the target area image and the search area image of the video to be processed, multi-dimensional target feature information is extracted, and object detection is performed based on these feature information to improve the accuracy of the detection.
Benefits of technology
It improves the accuracy of object detection and the positioning accuracy of tracking algorithms, and can more effectively handle target tracking tasks in complex scenarios.
✦ Generated by Eureka AI based on patent content.
Abstract
The present application relates to the field of image processing technology, and in particular to an object detection method, device, electronic device and storage medium for improving the accuracy of object detection. The method includes: obtaining a target area image corresponding to an initial video frame in a video to be detected, and a search area image corresponding to a video frame to be processed; extracting multi-dimensional target feature information corresponding to the target area image and the search area image respectively; obtaining an object detection result corresponding to the search area image based on the multi-dimensional target feature information corresponding to the target area image and the target feature information of the same dimension in the multi-dimensional target feature information corresponding to the search area image; and determining the target state information of the target object in the video frame to be processed based on the object detection result. Since the present application can simultaneously retain the spatial detail information and semantic information of the image by performing multi-dimensional feature extraction on the image, the accuracy of object detection is effectively improved.
Need to check novelty before this filing date? Find Prior Art