Motion detection method and device

Abstract

The invention discloses a motion detection method which comprises the following steps: establishing a background model of a video image by using a random filtering algorithm; updating the background model according to the input video image, and acquiring a corresponding binary foreground picture and an integrogram of the binary foreground picture; extracting motion areas in the integrogram by using integral rectangles; clustering the motion areas so as to obtain a motion target. The invention further discloses a motion detection device. By adopting the technical scheme, the accuracy in motion detection is improved, both the property and the speed are taken into account, higher calculation cost is needless, and the method and the device are particularly applicable to embedded type monitoring equipment.

Description

technical field [0001] The invention relates to the technical field of video image processing, in particular to a motion detection method and device. Background technique [0002] Motion Detection Technology, also known as motion detection, is one of the core technologies in the video surveillance industry. This function is available in mainstream video surveillance equipment. [0003] Currently, motion detection algorithms include motion vector method, background subtraction method and optical flow field method. [0004] The motion vector method is to judge whether there is a motion area in the image by extracting the motion vector of the macroblock in the video coding sequence. Its detection speed is fast, but it cannot accurately locate the position of the target, and it is easily affected by illumination changes, jitter, and noise. The false positive rate is extremely high. [0005] The background subtraction method is to set the motionless part of the video scene as ...

AI Technical Summary

Problems solved by technology

This method can obtain a complete moving target, but the traditional background modeling is sensitive to jitter and shadows, and the mathematical morphology analysis algorithm and connected domain analysis have high computational complexity, which will consume a lot of calculation and storage space. Embedded monitoring devices with limited computing power and storage space have high performance requirements

[0006] The optical flow field method is to analyze the moving area by calculating the moving field in the image. It has a huge amount of calculation, needs to be equipped with special hardware for acceleration, and is sensitive to noise, and is not suitable for embedded monitoring equipment.

Images