Research On Moving Object Detection Algorithm Based On STFD

As an important research content in video image processing,moving object detection technology has received much attention in recent years and is a hot research topic of many experts and scholars.As a key subject in the field of computer vision,moving object detection technology requires fast and accurate separation of moving objects in video sequence images from the background,and is the basis for subsequent object tracking,classification,recognition and behavior analysis.It is a research difficulty in image processing,three-dimensional reconstruction and other technologies.Therefore,the moving object detection algorithm with good real-time performance,fast calculation speed,high detection precision and good robustness is the key to our research.The following work has been done for the purpose:Firstly,it summarizes the background and significance of this research,the development and status quo and innovation of object detection technology,and expounds a series of implementation methods of video sequence reading and preprocessing,as well as the classical moving object detection algorithm.Secondly,in the implementation process of the frame difference method,"empty" and "double shadow" may cause incomplete and unclear detection object,and smoothing three-frame difference(STFD)is adopted.The image segmentation method and the kernel regression method are used to smooth the light sensitive area,and then a Gaussian mixture model with different cardinalities is proposed to propose a moving target detection algorithm.The positive and reverse order are used to process the sample gray matrix,and the detection is enhanced.Accuracy;the adjacent frame gray image is translated and rotated,and the overlapping part when the particles coincide;the affine invariant is derived,and the gray image recognition analysis is performed,thereby solving the influence of the camera shake onthe moving target extraction.The improved algorithm effectively overcomes the effects of camera shake or offset,as well as sudden changes in light and "ghosting",with high precision and robustness.At last,in order to extract the human moving object more accurately and efficiently in the surveillance video,a moving object detection algorithm combining smoothing frame difference method and Robust Principal Component Analysis(RPCA)is proposed.RPCA can achieve both data dimensionality reduction and high noise,spike noise rather than Gaussian distributed noise.The two algorithms are used in combination,and the background of the current frame of the RPCA extracted video is used as the intermediate frame of the smoothed frame difference method,and is respectively differentiated from the previous frame of the current frame and the current frame of the video,thereby avoiding the background pixel point.The influence eliminates the phenomenon of “emptiness” and also contributes greatly to the reduction of noise.Video detection experiments in different scenarios show that it is more efficient and accurate than similar algorithms.