Study On The Intelligent Monitoring System Of Moving Object

The thesis is focused on two emphases: study on the technology of detecting moving object; design the intelligent monitoring system which centers on the technology of detecting moving object.After summing up the defects of traditional monitoring systems, the merits, the wide application fields and the profound theories of intelligent monitoring system are analyzed. The technology of static image processing is the base of intelligent monitoring system, so this paper introduces several typical static image processing methods which used in this monitoring system. The technology of detecting moving object is discussed emphatically in this paper, the merits, the defects and the application fields of several motion detecting algorithms are compared first, then, according to the characteristics of this monitoring system, the optical flow algorithm is chose to detect moving object. The global motion estimation for compensation technology is introduced, and a new method of global motion estimation for compensation is brought forward, which combines the local areas method with the statistic method, the accuracy of global motion estimation is enhanced.At the part of this intelligent monitoring system design, first of all, the performance requirements, the hardware structure, the work sequence and the development tools are introduced respectively, then, the system is designed by two schemes. According to the characteristics of the schemes designed, to enhance the real time performance of this system, a new method is designed which can reduce the optical flow calculation by forecasting the area of detecting moving object according to the instantaneous velocity of moving object. In addition, according to the area forecasted, a new method is designed to select the local areas which are used for global motion estimation. Validated by experiments, the systemic designs achieve the performance requirements satisfactorily.