图像匹配方法、装置、存储介质及电子设备
By extracting feature points and contextual features of the target image during answer sheet review, and combining them with existing algorithms to obtain similar feature points, the matching error problem caused by similar texture features of the answer sheet is solved, and accurate recognition of the answer sheet content is achieved.
CN115546818BActive Publication Date: 2026-07-17BEIJING YOUZHUJU NETWORK TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING YOUZHUJU NETWORK TECH CO LTD
- Filing Date
- 2021-06-29
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing technologies often result in image matching errors during the answer sheet review process due to similar texture features on the answer sheets, leading to inaccurate identification of the answers and review errors.
Method used
By extracting feature points from the target image, the feature descriptor and context features of the first feature point are obtained. Similar feature points are obtained by combining them with a preset template image. The SIFT, SURF or ORB algorithms are used to extract and map feature points, thereby improving the accuracy of feature point extraction.
Benefits of technology
This improved the accuracy of image matching results, reduced the error rate in answer sheet review, and ensured the accurate recognition of answer content.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115546818B_ABST
Abstract
本公开涉及一种图像匹配方法、装置、存储介质及电子设备，该图像匹配方法通过对目标图像进行特征点提取，以得到该目标图像对应的多个第一特征点以及每个该第一特征点的第一特征描述子；获取每个该第一特征点对应的上下文特征；然后根据该第一特征点的该第一特征描述子和该上下文特征从预设模板图像中获取该第一特征点对应的相似特征点；将每个该第一特征点映射至该第一特征点对应的相似特征点上，以使该目标图像匹配至该预设模板图像，这样，能够结合该第一特征点的上下文特征确定该第一特征点对应的相似特征点，能够有效提升相似特征点的准确性，从而能够保证图像匹配结果的准确性，有利于降低答题卡审阅的出错率。
Need to check novelty before this filing date? Find Prior Art