The Reseach Of Image Edge On Detection Technology

Edge is the most basic and most important feature of the. Edge detection is always thecomputer vision and image processing field the most classic topic. Image analysis andunderstanding of the first step is often to edge detection. The goal of edge detection is to findthe image about shapes and reflectance or transmittance is information, image processing,image analysis, pattern recognition, computer vision and human vision is one of thefundamental steps, the correctness and reliability of its results will directly affect the machinevision system to the objective understanding of the world. However, with the development ofmodern science and technology and modern industry of microfabrication technology demand,the traditional edge detection technology can not meet the requirements of measurementprecision. Sub pixel subdivision technology is the demand of a new technology, it is in highprecision in image measuring system using greatly improves the precision of measurement.In this paper the concept and algorithm of edge detection based on edge detectiontechnology, summarized the development present situation, introduced several classical edgedetection algorithms, and presents a new method based on human visual threshold Kirschedge detection method. The traditional edge detection operators are generally only get verywide edges for the subsequent image processing, this brought certain difficulty. Based on theimage edge feature analysis, classical refinement method, this paper proposes an improvedZhang and Suen thinning algorithm will get the image edge thinning. The simulation resultsshow, the method to get the accurate edge, good continuity.At the same time in order to adapt to the increasingly high demands on the edge of imageprocessing, need further sub-pixel processing, so the sub pixel detection was introduced to thethinning process. First discusses the subpixel edge location of origin, principle and premisecondition, summarizes the existing subpixel edge detection method, obtains the Zernikemoment method is a good sub pixel detection method. Finally, Zernike moments to be thinnededge image sub-pixel positioning, through the simulation experiments were obtained after thesub-pixel image edge.