Characteristic Detecting And Its Application In Object Recognising

Computer Vision has been developing as a multidisciplinary subject. Its containsvarious field needed to be researched.area spread widely.It is meanful to study thebottom layer of Marr'CV theory, also, Studying on characteristic detecting which lies inbottom layer of Marr'CV theory has important significance not only on science butalso on practical application.In the paper, how to recognize the chinese chess in the chess broad isdiscussed ,SIFT characteristic vector is used as the primary way in chess recognizingand color and logic is assistant. Finally, a good result is reached.The main content of this paper is listed here:(1)description of computer vision'sdevelopment and its application;(2)studying on wide-baseline correspondence andshort baseline correspondence matching; and some comer point detectingmethod;(3)studying on the theory of SIFT and how to use it, also, the experiment resultis showed;(5)the arithmetic of figure-process is chosen by simulation which providedtheoretics base;(6)description of the frame of the system and some code is provided.A new method of object recognize is given here,and it can be used inrobot-football and object tracing and etc.