The Study And Application Of Edge Extraction In Image Matching

Image matching is widely used in advanced technology, especially a core tech at terminal guiding processing of cruse missile. So a competition can not be avoided .Through it made a obvious progress at the near 20 years, lots of problems still deserving a solution. Laplace operator edge extraction is a key step between character getting and matching. A less complicated calculating made it more effective. Unlucky, a much noise result too. A rational optimize will make it fit for the application which demand more effectiveness. The paper summed up the problems. It analyzed the possibility of transplanting a new algorithm called pixel runs labeling algorithm serving a dimensional structure filter for Laplace operator edge extraction. Based on the pixel runs labeling algorithm the improved algorithm with dimensional structure filter overcomes the disadvantage of noise without lowering the speed greatly. The algorithm put forwarded in this paper is quite practical in the application that highly needs real-time processing.