This paper presented an image segmentation method based on the symmetric feature. We cut a butterfly out of the background using image-processing technique, which employs the characteristic of the symmetric in shape and stripe of butterflies. We use the K-mean method to classify targets by color, what we want to do is divided an image into several regions with similar color, and then we cut symmetric objects in the image by judge the symmetric of each regions. From the result of our experiment, our proposal can cut butterflies in lots of images.
- 非天使Lv 71 decade agoFavorite Answer
This research presents a segmentation method that features symmetry, we take advantage of the characteristic of symmetry in shape and stripe of a butterfly to cut its image out of a picture’s background with the image-processing technique, in the beginning ,we use the K-mean method to classify the colors of the target image , what we want to do is dividing the image into several regions which of similar color, then we cut symmetric objects out of the image by recognizing the symmetry of each region, and depending on the result of our experiment, it shows that our proposal can exactly cuts out butterflies which in the images .
2008-11-10 19:34:36 補充：
....................................., we take advantage of the characteristic of symmetry in shape and stripe of a butterfly and the image-processing technique to cut the butterfly's image out of a picture's background,