The Realization Of The Moving Object Detection Algorithm In Stale Scene

The moving objects detection is one of the hottest researches in the field of video process and analysis. By using moving objects detection, the moving objects are separated from complex background, and the description of objects and their movement parameters are obtained. The moving objects detection is the basis and key to solve many computer vision problems, and the process of objects classification, tracking and behavior understanding is affected by the results of the moving objects detection. At present, the moving objects detection has been widely used in the areas of image processing, industrial inspection and Intelligent System, but there are many deficiencies about it on the theory and application. A lot of researches have been done and many results have been achieved by domestic and foreign scholars. An improved algorithm of moving objects detection based on Gaussian mixture model is proposed in this paper based on the results of the researches, and the algorithm has been achieved on the hardware platform of which the core is TMS320DM642.Firstly, the research of moving objects detection algorithms has been done. The three main objects detection algorithms including consecutive frames difference, optical flow and background subtraction are have been analyzed and compared. An algorithm for moving objects detection based on Gaussian mixture model and three-frame differencing is proposed to make up the deficiency of the background subtraction method based on Gaussian mixture model. The simulation of the proposed algorithm has been carried out by MATLAB, and the simulation results indicate that the proposed method could extract the moving objects effectly and possess faster processing speed and better effect.Secondly, a moving object detection system is designed based on DSP, and the detailed design of the hardware interface between the module and the bus connection is given. The video capture module, the DSP module, the video display module and the memory module are included in the system. The TVP5150PBS is used as decoder for the video capture module, the TMS320DM642 is used as core computing processors for the DSP module, the SAA7121H is used as encoder for the video display module, the HY57V283220TI is used as SDRAM and the MBM29DL800TA is used as Flash for memory module.Thirdly, this paper has carried on feasibility research on how to implement the moving object detection algorithm on the hardware system. The flow chart of the system software is designed arid the writing of the system software code is completed. From aspects of C code and data storage, the size of the code is reduced and the running speed of the algorithm is increased by optimizing the system software.Finally, the system is analyzed and tested. Through the test of the basic system module and the video processing module, the effect of the algorithms is verified.