Research On Moving Target Detection And Tracking Algorithm In Video

Target detection and tracking is an important part in the field of computer vision.In recent years,due to the development of the Internet,big data,and cloud computing,vast amounts of data have been provided for computer vision.The innovation of the algorithm and the major breakthrough in computing power have made significant progress in the image processing,which provides a solid foundation for the development of artificial intelligence.In this paper,vehicle video is taken as the research target,and the target in the video is detected by the convolutional neural network algorithm.A system platform specially designed for traffic monitoring is designed.The major work completed this time is as follows:1.Research and analysis of target recognition algorithm.In this paper,the neural network and deep learning algorithms are used to classify and identify the target,and the neural network based on logistic regression is used to classify the face and nonface of the image.Convolution neural network is used to identify multiple targets such as dogs,birds,pedestrians,automobiles and insects.And through experiments,relevant algorithm research is analyzed and optimized.2.Research and analysis of target location algorithm.The traditional target location algorithm,such as sliding window and selective search,is introduced.In order to reduce computation redundancy and increase robustness,the target location algorithm using sliding window in convolution neural network is expounded.The advantages and disadvantages of existing deep learning network models for target localization are also introduced.3.The realization of the target detection system.Car video processing,data collection and offline training convolutional neural network model,build a software platform in MyEclipse,using deep learning for target detection,and design a set of intelligent traffic monitoring system to meet the actual needs.