Research Of Recognition Of Pointer Image In Video Surveillance System

Image processing is the key question which is concerned with extensive fields in recent years. This thesis summarizes the present condition of the digital image processing technology, combining the application of the project, emphatically discusses the method of processing and recognition of pointer image in digital video surveillance system.VC++ is applied as the developing tools in the video surveillance system. 24bit-real-color picture to 8bit-gray-scale picture, threshold, image dilation, image thinning are applied to carry through image pretreatment; two kinds of feasible scheme of image recognition are put to use: area and angle recognition method; in the area recognition method, point seeking and warning area pointer seeking algorithms are contrasted and concluded. Angel recognition method is selected in the experiment, designs and tests the recognition of the pointer image. At last, the angel recognition method is proved to be all right and satisfies the requirement of the system.