Research On Moving Object Detection And Tracking Algorithm In Video Surveillance

With the development of moving target detection and tracking technology,as well as the scholars research of the compressed sensing theory.The background modeling and compression show great value in the theroy research and engineering applications,especially to meet the precision,speed.However,due to the complexity of the background,there is always have noise and the target occlusion,such problems show that the compression and tracking system is still needed to improve performance.In this paper,the mixture gaussian background model and the compressive tracking algorithm have been studied based on research of predecessor,the mixture gaussian background model have been optimized and the compression tracking algorithm was improved.In order to speed up the algorithm operation speed of mixture gaussian model,this paper proposed a mixture gaussian model based on improved neighborhood mean.Consider the calculated data redundancy when establish the mixture gaussian model for background,use 3*3 templates for video frame pixels and calculate the mean plus factors are poor,then establish mixture gaussian model based improved Neighborhood mean.By comparing the algorithm execution time experimental results of the conventional pixel gaussian mixture model and the improved model,the results show that the improved method of this paper has better robustness,and at the same target effect of the premise,based on improved algorithm execution rate mean gaussian mixture model to perform up to three times the rate of conventional pixel algorithms.In the study of compression tracking algorithm,the compressive tracking algorithm in the tracking process can cause the failure when the target moves fast and has big changes in appearance.In order to solve these problems,a developed compressive tracking algorithm has been proposed.Firstly,the sample sets have been improved,excepting the positive and negative sample sets,we have also collected one additional candidate sample set,which can effectively reduce the error of gathering samples.Secondly,referencing the MIL algorithms,strong classifier is constituted by the assembled weak classifiers,and it will determine the target through calculating the candidate sample sets,we also propose new methods which can update the learning rate and the strong classifier.Finally,we use the probability values that the candidate sampleset is positive sample set to adaptively adjust the size of the target window.Experiments show that the improved algorithm has an accurately tracking result for the fast moving and big changed targets,and it also has better robustness than the other algorithms.