一种双录视频相似分区智能检测方法及系统
By combining image processing and feature matching technologies with a similar partitioning intelligent detection method for dual-recording videos, the problem of accuracy in identifying primary and secondary partitions in dual-recording videos is solved, achieving efficient and accurate partitioning detection, which is suitable for diverse dual-recording scenarios in financial institutions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGDONG MICROPATTERN SOFTWARE CO LTD
- Filing Date
- 2025-12-18
- Publication Date
- 2026-07-17
AI Technical Summary
Existing quality inspection and auditing systems struggle to efficiently and accurately identify primary and secondary sections with similar content when processing dual-recorded videos. This is especially true when the primary section includes dynamic TTS subtitles and compliance prompts, which affects the accuracy of feature extraction and leads to a high false positive rate.
A dual-recording video similarity partitioning intelligent detection method is adopted, including parsing the video GOP structure, adaptive line detection, dynamic rectangle generation, accurate sub-image segmentation, three-level classification model and feature matching technology, combined with image processing, text erasure and key point matching, to achieve accurate detection of primary and secondary partitions.
It significantly reduces the false negative rate and improves the accuracy and efficiency of primary and secondary partition detection in dual-recording videos. It is suitable for diverse dual-recording scenarios, especially those that require capturing dynamic interactive actions, such as signing processes and document presentation.
Smart Images
Figure CN121661569B_ABST