Research On Video Moving Target Detection Algorithm Based On Smart Mobile Terminal

Moving target detection technology is the basis of the field of computer vision.This technology mainly extracts interested targets from video according to the difference of pixel changes between the background area and the moving area in video.The quality of target detection will directly affect the follow-up tracking,monitoring,abnormal behavior detection and other purposes.The video captured by traditional fixed cameras mostly in a static background,but in recent years,with the popularity of smart phones,aerial cameras and other mobile devices,people capture more video information through smart mobile devices year by year.Direct detection of moving targets in video through smart terminals can reduce resource consumption.However,due to the influence of the environment and the movement of the device during the capturing,these video backgrounds are no longer static.Therefore,in view of the above problems,this paper comprehensively considered the hardware limitations and memory consumption of intelligent terminals,combined with the background compensation algorithm,designed and realized the video moving target detection algorithm based on smart mobile terminals.The main research contents of this paper are as follows:1.For the background offset caused by camera device offset,traditional background compensation algorithm was studied and analyzed,and a background compensation algorithm based on improved SURF algorithm is proposed.First SURF algorithm is used to extract features of detected images,and binary feature description algorithm is used to describe feature points.Then hamming distance is used to match and RANSAC algorithm is used to eliminate mismatched points.Compared with the traditional algorithm,the improved algorithm has higher accuracy and robustness while improving real-time performance.Experimental results show that the proposed algorithm has good matching effect under different scales,rotation angles and illumination intensity.2.Aiming at the hardware limitations of smart mobile terminals and the problem of dynamic background detection,this paper proposes an improved Vibe algorithm combined with background compensation algorithm.First background model is established according to the first frame of video,then each frame of video sequence is matched and compensated according to the background compensation algorithm and the updated model,so the background of each frame of video is kept relatively static with the mobile device.At the same time,the updating factor of Vibe method is kept constant,and the target detection effect was improved by combining the edge detection algorithm and morphology.Experimental results show that the proposed algorithm performs well in the dynamic background.3.Aiming at the application research of target detection on smart mobile,background compensation and video moving target detection system based on smart mobile terminal are designed and implemented to meet the needs of mobile intelligent monitoring.This system is based on the Open CV library and Android development platform and combined with the background compensation and target detection algorithm proposed in this paper.In addition to the advantages of the above algorithm,the system has high efficiency in background compensation and target detection,good detection effect in mobile state,and low development cost.