| Summary: | 碩士 === 義守大學 === 資訊工程學系碩士班 === 93 === Active contour model (ACM) is the technology that a kind of image segmentation, used for detection of object in the image mainly, and have obvious treatment results to high noised and low contrast image. Its method is based on initial control points and image characteristic, ex. gradient, it is internal energy and external energy to define, which find optimal solution by way of concept of Euler-Lagrange equation, approach through conservation of object contour. But the traditional ACM has two great disadvantage, first, the problem of set up initial control point, second conserve the hollow of the object, these two disadvantages often make the result not good finally, and need to spend more time on operation. Particle swarm optimization (PSO) is a kind of evolution computing method, calculation is very simple and fast in operation, for looking for the best solving of some function or searching to a range, the result is very good. So this thesis proposes using the PSO to strengthen the result of minimization of ACM energy function, and improve the influence of detection of object for gradient value, help the ACM to solve the above two disadvantages, have optimized the result of conserving.
|
|---|
