Abstract
With the development of science and technology, home monitoring security system has gradually become an important means for the family. Home monitoring system to improve more and more attention by the community. This article has designed a family-based intelligent motion detection surveillance system designed to combine home monitoring several characteristics: small, high precision, remote monitoring, the production of low cost, applicable to public family.
Monitoring in the home moving target detection is based on the bottom of the video and more information on testing, as the majority of image processing is taken into account the corresponding movement of the pixel region, the regional campaign for the effective extraction of the target classification, target tracking as well as the work of understanding importance. Moving target detection has extremely broad range of applications, such as: military reconnaissance, intelligence surveillance and testing in areas such as transport. For moving target detection in image sequences can be divided into a static background, the background of both cases movement, this article is based on the movement of the background research.
First, an overview of moving target detection of several commonly used methods, including optical flow method, inter-frame difference method, the background reduction method, on the various methods were compared and analyzed the advantages and disadvantages. On this basis this paper, a static background of the moving target based on boundary detection algorithm, the algorithm for computing the first two images adjacent to the closure of the value of difference image, the difference image into a binary image, and then to the binary image to form filter learning to identify the rough location of moving targets, while the current frame edge detection, edge detection results to determine the movement of the former position to pay target rough calculation, to be moving targets in the border, and then by the filtering, filling, such as regional processing, the final moving target access area.
Keywords:Image acquisition, motion detection