Image Deformation Based On Contour

Image deformation is widely researched this years, it’s the most important leaf ofimage processing. Because of its wide usage, it has abstracted much attention. Thealgorithms used now is discovered in many image processing software, the uses coulddrag the points on the picture to paralyze an image.The algorithms now would intersect a series of feature points or feature line in theimage. The user could drag them to drive any other points to move towards them. Anyminor movement will bring a change on the picture. These feature points or featureline’s positions will restrict the points’ location nearby and drive them to move.Although the movement will obtain a contact a picture, it will affect the pixels on thepicture which is not in the deformation zone. So, it’s can’t be widely accepted as aresult of the defects.In order to treat with the defects of the algorithms nowadays and make the imagedeformation more widely used, we propose an image deformation algorithm based oncontour, because of the contour could make a segment on the origin pictures, thedeformation operation can’t have any effect on the other zone outside the deformationzone. So the algorithm will get rid of the obvious defects of the image deformationresults. In addition, the uses’ operation is based on contour, so they can deform thecontour to get the shape they want.When the users get the shape they want, they also get the deformation zone. Andthe pixels in the zone have not been imputed yet. We then use the min least square torestrict the deformation zone’s pixels to move to keep the features and textures in thedeformation zone.It can also keep the smoothness between the deformation zone andthe non-deformation zone.The result has proved that this algorithm can paralyze an image properly,as wellas get an wonderful color and texture feature. Because of the algorithm keep thenon-deformation zone unchanged, so it has a bright future.