Monitoring video processing method and apparatus

By constructing a coupled analysis of temporal error diffusion gradient parameters and frequency domain energy dispersion parameters, and dynamically adjusting the reference frame set, the error diffusion problem of surveillance video under low bitrate or long GOP conditions is solved, improving video stability and detail preservation capabilities.

CN122269032APending Publication Date: 2026-06-23SHANGHAI LINGJIN BIG DATA TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANGHAI LINGJIN BIG DATA TECHNOLOGY CO LTD
Filing Date
2026-04-28
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Existing video surveillance processing methods, under low bitrate or long GOP conditions, are prone to local distortion spreading into the time domain due to transformation and quantization errors, resulting in frame-by-frame degradation, flickering, ghosting, and structural distortion, lacking effective means of identification and suppression.

Method used

By constructing a dual-feature analysis mechanism of temporal error propagation gradient parameters and frequency domain energy dispersion parameters, an error propagation intensity index is generated, the reference frame set is dynamically adjusted, the weight of distant reference frames is reduced, and residual data reconstruction processing is performed to suppress error propagation.

Benefits of technology

It significantly improves the sensitivity and stability of distortion detection, reduces the probability of false positives and false negatives, and ensures the stability and detail retention of surveillance video.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122269032A_ABST
    Figure CN122269032A_ABST
Patent Text Reader

Abstract

This invention discloses a method and apparatus for processing surveillance video, belonging to the field of video processing technology. First, predictive coding and residual transformation are performed on the input video sequence to obtain reconstructed data and residual frequency domain coefficients. Then, temporal error propagation gradient parameters are extracted based on the reconstructed data, and frequency domain energy dispersion parameters are extracted based on the residual frequency domain coefficients. Further, the two types of parameters are normalized and coupled to generate an error propagation intensity index, which is used to determine whether there is a risk of distortion propagation. When a risk exists, high-risk coding regions are located, the reference frame set is dynamically adjusted, and the weight of distant reference frames is reduced. Simultaneously, the residual data is reconstructed, and the reconstructed data is compensated and updated. Through the above technical solution, effective suppression of video coding error propagation is achieved, significantly improving the stability and quality of surveillance video under low bitrate or long prediction structure conditions.
Need to check novelty before this filing date? Find Prior Art