Robust foreground detection method based on multi-view learning

Abstract

The robust foreground detection method based on multi-view learning provided by the present invention includes: obtaining a reference background image through the time-domain median filtering method of the input video, performing iterative search and multi-scale fusion acquisition on the current image and the reference background image Heterogeneous features; use the conditional independence of the heterogeneous features to calculate the conditional probability density of the foreground class and the conditional probability density of the background class, and use Bayes' rule to calculate the posterior of the foreground according to the foreground likelihood, background likelihood and prior probability The posterior probability of probability and background; construct the energy function of Markov random field model through the posterior probability of described foreground, the posterior probability of described background and space-time consistency constraint, utilize belief propagation algorithm to minimize described energy function Segmentation results of foreground and background are obtained. The invention can realize robust foreground detection in complex and challenging environments.

Description

technical field [0001] The invention relates to intelligent video monitoring technology, in particular to a robust foreground detection method based on multi-view learning. Background technique [0002] Intelligent video surveillance is an important means of information collection, and foreground detection or background subtraction is a very challenging underlying problem in intelligent video surveillance research. On the basis of foreground detection, other applications such as target tracking, recognition, and anomaly detection can be realized. The basic principle of foreground detection is to compare the current image of a video scene with a background model and detect regions with significant differences. Although seemingly simple, foreground detection often encounters three challenges in practical applications: moving shadows, illumination changes, and image noise. Motion shadows are caused by light sources being occluded by foreground objects, which are hard shadows ...

AI Technical Summary

Problems solved by technology

Second, only the background model is established, and the foreground pixels are identified as outliers, and it is difficult to distinguish the foreground with similar color to the background

Third, no spatio-temporal consistency constraints in video sequences are exploited

Images