Moving object detection method with combination of sample consistency and local binary pattern

Abstract

The present invention proposes a moving object detection method with the combination of sample consistency and a local binary pattern. A background model is established for each pixel of a video imagesequence in each frame of the image, the differential operation of a current image after an Nth frame of image of a visible light video sequence and a previous frame of image is carried out, images which are subjected to differential operation are grayed and binarized, a stable background point is determined, and a target foreground is extracted. Through a flooding filling algorithm and connected-domain detection, a large number of voids is filled and discrete noises are removed, the concept of the stable background point is proposed, and a foreground target is accurately judged in the case of a sudden change in light with the combination with the robustness characteristics of the LBP (Local Binary Pattern) for illumination change.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a moving target detection method combining sample consistency and local binary mode. Background technique [0002] Moving object detection is an important research topic in the field of computer vision, and has a wide range of applications in many fields such as intelligent video surveillance, human-computer interaction, and visual navigation; it is at the bottom of the entire video processing system and is a variety of subsequent advanced processing, such as The basis for object tracking, object classification, behavior recognition, and scene understanding. [0003] At present, the commonly used moving target detection methods mainly include the following three methods: optical flow method, frame difference method and background modeling method. [0004] (1) The optical flow method defines the instantaneous rate of change of the gray scale on the coordinate points of t...

AI Technical Summary

Problems solved by technology

The SACON algorithm has fast operation speed, strong anti-noise ability, and strong background adaptive ability; however, because it extracts possible motion foreground through neighborhood frame difference, even if some holes are filled in the later stage, the extracted motion There will still be some holes inside the target; in addition, the SACON algorithm uses the first N frames of the video sequence as the sample set. When the illumination changes slowly or suddenly changes in the subsequent video, the sample set is too late to update, which will cause the illumination change area to be judged as Foreground area, which leads to misjudgment of moving targets

Images