The implementation of using P-P-C filter to balloon snake algorithm

Using balloon snake algorithm to get the contour in an image, the contour will stay on the inner darker place (that is in the relatively minimum energy) of the image. To solve this problem, we usually use median filter or Gaussian filter as the tool of softening brightness of the image. These two filters will blur the edges of the object, thus increase the difficulty to find the real contour of the object. Papari, Petkov and Campisi propose an algorithm (P-P-C filter) to soften the internal part of an image and strengthen the edge of the objects in the image. When we use balloon snake algorithm to process an image which relative contrast is not obvious, using P-P-C filter as the pre-process tool is still easier to identify the contour of the object of image than using median filter and Gaussian filter. In this paper, we use an example to illustrate that using P-P-C filter then use balloon snake algorithm to find the contour of an object is better than direct use balloon snake algorithm, using median filter then use balloon snake algorithm, and using Gaussian filter then use balloon snake algorithm to find the contour of an object.