A method and apparatus for monitoring video storage, an electronic device and a storage medium

CN123035280APending Publication Date: 2026-11-03SHANGHAI ZHIWEI ROBOT CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611151749.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-31
Publication Date
2026-11-03

AI Technical Summary

Technical Problem

Existing surveillance video storage systems cannot differentiate storage for different situations, predictive output lacks confidence information, fusion weights are static and do not consider the confidence quality of the prediction itself, storage decisions rely solely on the absolute value of density rather than confidence, do not consider confidence dimensions, do not classify sources of uncertainty, and adopt a one-size-fits-all approach to feedback calibration.

Method used

By acquiring behavioral signals from home surveillance videos to be stored, historical model parameters and home environment context information are determined. A pre-trained behavior evaluation model is used to generate behavior scores. Storage evaluation parameters are generated by combining video information with uncertainty and semantic density information, and differentiated video storage strategies are matched.

Benefits of technology

This enables differentiated storage for surveillance video storage systems, improves the accuracy of storage strategy adaptation, avoids a single evaluation dimension, and enhances the credibility and adaptability of storage decisions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123035280A_ABST
    Figure CN123035280A_ABST
Patent Text Reader

Abstract

The application discloses a kind of monitoring video storage method, device, electronic equipment and storage medium.The method comprises: obtaining the behavior signal of the home monitoring video to be stored, and determining historical model parameters and home environment context information according to the home monitoring video;According to the behavior evaluation model pre-trained, the behavior score of behavior signal is determined;According to the data fluctuation degree of behavior score, historical model parameters and home environment context information, video information comprehensive uncertainty is generated;The semantic density information of the home monitoring video is obtained, and storage evaluation parameters are generated according to video information comprehensive uncertainty and semantic density information;Based on the video storage strategy matched by storage evaluation parameters, the home monitoring video is stored.The technical scheme of the embodiment of the present application can avoid the single evaluation dimension by generating video information comprehensive uncertainty;By storing video using video storage strategy, the rationality of storage mode can be improved.
Need to check novelty before this filing date? Find Prior Art