Object recognition method, electronic device, and storage medium

CN122289735APending Publication Date: 2026-06-26ZHEJIANG DAHUA TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
ZHEJIANG DAHUA TECH CO LTD
Filing Date
2026-03-10
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

In the process of object recognition, the accuracy of object recognition results decreases due to occlusion of the target object in the image to be recognized or low image quality. In particular, when the similarity between different objects is high but the similarity between the same object is low, existing technologies are difficult to effectively improve the recognition accuracy.

Method used

By performing feature extraction and masking image processing on the image to be identified and the image pairs in the image database, the target features of the image to be identified and the database images are generated using an object recognition model. Combined with feature similarity calculation, the effective regions in the image are automatically focused to improve the accuracy of feature similarity, thereby improving the accuracy of object recognition results.

Benefits of technology

By guiding the object recognition model to focus on the effective regions of the image, the similarity of target features between the image to be recognized and the base database images is improved, thereby increasing the accuracy of the object recognition results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122289735A_ABST
    Figure CN122289735A_ABST
Patent Text Reader

Abstract

This application discloses an object recognition method, an electronic device, and a storage medium. The object recognition method includes: forming at least two image pairs to be recognized based on an acquired image to be recognized and images in an image database; inputting the image to be recognized, a first mask image of the image to be recognized, and the corresponding database image into an object recognition model to obtain a first target feature of the image to be recognized; inputting the database image, a second mask image of the database image, and the image to be recognized into the object recognition model to obtain a second target feature of the database image; and determining the object recognition result of the image to be recognized based on the feature similarity between the first target feature of the image to be recognized and the second target feature of the database image. This method can improve the accuracy of object recognition results.
Need to check novelty before this filing date? Find Prior Art