A large-scale near-duplicate video retrieval method based on random multi-view hashing
A large-scale, near-repetitive technology, applied in the direction of video data retrieval, video data query, special data processing applications, etc., can solve problems such as fast speed, retrieval accuracy and scalability limitations, and retrieval speed cannot be satisfied, and achieve real-time The effect of retrieval
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0038] The method in the present embodiment is first to extract video keyframes by the method of time interval sampling, and extract the color histogram HSV feature and local binary pattern LBP feature of keyframe; Then, by linear mapping and sigmoid function, color histogram HSV features and local binary pattern LBP features are mapped to the quasi-hash space, and the hash code generated by thresholding to obtain the final representative video hash code; finally, the key points are calculated in the feature space and the quasi-hash space respectively. Gaussian conditional probability between frames, using a composite Kullback-Leibler (KL) divergence to measure the consistency of two conditional probability models, and using a standard gradient descent method to optimize the combination coefficient and bias parameters of the hash function . After the parameters of the hash function are trained, each video will be represented by a string of binary hash codes, and a fast Hamming...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com