The Moving Object Detection And Tracking Algorithms For Outdoor Video Surveillance

With the widely application of outdoor video surveillance system, more and more surveillance videos have been produced. One has to manage and real-time mine information and knowledge of interest from the large-scale videos, in order to realize intelligent video surveillance. The moving object detection and tracking methods are the most basic and important technology in the area of intelligent video surveillance, and are the key to realizing real-time intelligent video surveillance.This paper presents a set of moving objects detection and tracking algorithm, can carries on the stably tracking in the complex circumstance to the single-object and multi-objects.The paper's main achievements include following several aspects:A background reconstruction algorithm used in background subtraction motion objects detection is proposed.The algorithm integrates the stabilization duration with appearance frequency,and uses weighted intensity histogram to confirm the background probability of quantity interval.Based on the probability values and distributing realizes the background reconstruction.Then a cleaner background can be extracted from the background video sequence with moving target and the movement ghost also be eliminated with the algorithm.A fast motion detection on dynamic background algorithm is proposed ,which is detecting moving target fast using an adaptive local threshold. The adaptive local threshold is created with the standard deviation of difference image region entropy ,The algorithm can effectively suppress the background disturbance and detect the small target in undisturbed region.A based on predicted and features motion objects tracking algorithm is proposed. It predicted the target location by Kalman filter to reduce the search range of motion objects matching, and speed up object matching by hierarchical matching algorithm with double layer target feature set. The simple features in the double layer target feature set are used for handling with the most case such as tracking the vehicles and pedestrians moving in line. The random edge feature, a complex feature of the double layer target feature set, is selected to track the temporary occlusion of vehicles stability. Ultimately, we can track the vehicles and pedestrians in real-time and stability on the outdoor video surveillance.This algorithm has been verified by multiple of video surveillance for Image sequences and the results are accurate, which prove the good robustness of this approach, the experiment results are satisfied as well.