Stranger recognition method and device, electronic equipment and storage medium
By using a combination of segmentation algorithms and saliency detection in pedestrian recognition, the problem of low accuracy in pedestrian recognition in existing technologies is solved, and accurate identification and elimination of pedestrians and objects are achieved.
CN122176747APending Publication Date: 2026-06-09BEIJING OPPO TELECOMM CORP LTD
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING OPPO TELECOMM CORP LTD
- Filing Date
- 2024-12-06
- Publication Date
- 2026-06-09
AI Technical Summary
Technical Problem
In existing technologies, the accuracy of pedestrian recognition is low, especially in the detection of dense human figures and shadows, where there are problems of false recognition and missed recognition.
Method used
A segmentation algorithm is used to identify all targets in the image. Combined with saliency detection, the mask images of people and objects are extracted and combined to form a passerby image, thereby improving the recognition accuracy.
Benefits of technology
By combining segmentation algorithms and saliency detection, pedestrians can be accurately identified and eliminated, avoiding false and missed identifications, and improving the comprehensiveness and accuracy of pedestrian identification.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122176747A_ABST
Abstract
This application discloses a method, apparatus, electronic device, and storage medium for pedestrian recognition, relating to the field of image processing technology. The method includes: determining an image to be recognized; obtaining a segmentation result image of the image to be recognized and category labels corresponding to the segmentation result image; extracting a human face mask and an object mask from the image to be recognized based on the segmentation result image and category labels; performing saliency detection on the image to be recognized; identifying the main subject and pedestrians from the image to be recognized; extracting the human face mask corresponding to the pedestrian from the human face mask; extracting the object mask corresponding to the pedestrian from the object mask; and combining the human face mask and the object mask corresponding to the pedestrian to obtain a pedestrian image. This application, for the image to be recognized, uses a segmentation algorithm as a preprocessing step to identify all targets in the image, and the subsequent combination with saliency detection can achieve the purpose of intelligent pedestrian recognition, thereby improving the accuracy of pedestrian recognition.
Need to check novelty before this filing date? Find Prior Art