Research On Suppression Algorithm Of Foreground Static Area And Shadow Influence In Moving Target Detection

The purpose of moving target detection is to extract meaningful moving targets from video sequence images.Moving target detection is an important research content in the field of computer vision.In recent years,it has been widely applied in many fields,such as video surveillance systems,radar remote sensing images and unmanned driving.The accuracy of moving target detection will directly affect both the analysis and the understanding of the target's subsequent behavior.It is studied that how to improve the accuracy and integrity of the detected moving targets under several interference conditions when the detecting moving targets based on the background subtraction in this dissertation.To solve the problem of background model change which is caused by the sudden stop of the moving target in the monitoring scene and the start of the motion of the stationary object in the original scene,a new background updating algorithm that is based on the detection of the foreground stationary region is proposed to improve the accuracy of the background model in this dissertation.First,the non-moving pixel points existing is determined in the foreground.Secondly,the stationary region in the foreground is searched according to the connected region,and the stationary region that meets the condition is determined by calculating the area and the average value of gray pixels about the connected region.Finally,the background pixel is replaced by the corresponding pixel of the foreground stationary region to complete the update of the background model.The experimental results show that the background update algorithm is proposed in this dissertation can update the background of the foreground stationary region accurately and quickly,so as to achieve the accurate extraction of moving targets.To solve the interference problem of moving target shadow in grayscale video sequence images,a shadow removal algorithm of the four-direction gradient that is based on the grayscale ratio is proposed in this dissertation.Considering the shadow detection is affected by the change region of the grayscale ratio about moving targets is similar to the shadows,and the correlation between the shadow and the background is much higher than that between the moving target and the background,Normalized Cross-correlation(NCC)feature is combined to remove shadow and retain moving target in this dissertation,so as to ensure the integrity of the moving target.The qualitative and quantitative analysis of experimental results show that the algorithm presented in this dissertation is superior to other shadow removal algorithms in shadow removal ratio and keeping the integrity of the detected target.To solve the interference problem of the moving target shadow in color video sequence images,a shadow removal algorithm of the four-direction gradient that is based on the brightness ratio is proposed in this dissertation.Considering the shadow detection is affected by the change region of the brightness ratio about the moving targets is similar to the shadows,and the difference of the color feature between the shadow and the background is much lower than that between the moving target and the background,the color feature is combined to remove the shadow and retain the moving target in this dissertation,so as to ensure the integrity of the moving target.The qualitative and quantitative analysis of experimental results show that the algorithm presented in this dissertation is superior to other shadow removal algorithms in shadow removal ratio and keeping the integrity of the detected target.Through the research on these different interferences in the moving target detection,the moving target detection methods which are proposed to suppress these interferences in this dissertation.The experimental results show that the proposed algorithms can improve the accuracy and integrity of detecting moving targets.