Design And Implementation Of Moving Target Detection And Tracking System

In recent years,with the rapid development of the Internet,the data information is also growing rapidly,and computer vision research is also booming.The motion target detection and tracking have always been important research content in computer vision.The motion target detection and tracking are applied in the fields of intelligent security,intelligent transportation,urban planning,national defense and military field.Many scholars have been committed to do the research of target detection and tracking,but there are still a lot of problems that have not been solved yet,so the research of target detection and tracking is also of great significance and value to the development of computer vision.The main work of this paper is as follows:1)Research on image preprocessing technology,and the image preprocessing methods such as mean filter,median filter and morphological operation are discussed.Then the three methods of target detection are studied.The first one is the background difference method.The establishment of the background model uses the mixed Gaussian model and the KNN model is used for the experiment and the contrast analysis is carried out.The second is optical flow detection,the sparse optical flow and the dense optical flow are respectively tested and analyzed,the third is color detection.Finally,the three detection algorithms are compared again,and the advantages and disadvantages of each algorithm and the applicable scenarios are explained.2)The traditional method of target tracking is studied and improved.First of all,the Camshift algorithm is introduced.When the target is in a complex background and has the same color interference or is seriously occluded,the Camshift algorithm can not track well.Secondly,the Camshift algorithm is combined with the Kalman algorithm.When the tracking target is disturbed or severely occluded,the Camshift algorithm is no longer used.Because the Camshift algorithm is disabled,the Kalman filter is not updated using the optimal target position calculated by Camshift as an observation value,but the predicted value calculated using the Kalman filter is updated as an observation value.Finally,the improved algorithm is tested and the data is given to compare and verify that the improved algorithm has higher tracking accuracy and better robustness.3)This paper completes the design of the five modules of the system: video acquisition module,video preprocessing module,moving object detection module,moving object tracking module and browsing module,so as to realize the detection and tracking of moving objects in the video sequence,and can also browse and view the coordinates and frames of the tracking objects.