Image Deformation Using Least Absolute Deviation

Image deformation is a technology that produces fluid transformation of image A to image B.Morphing algorithm is actually a mathematical transformation between source and target images, which consists of two processes:warping and color interpolation. In the thesis we discuss the principle and steps of 2D image morphing.The thesis introduces several fundamental computer graphics algorithms and graphics deformation,and focus on an Image geometric Deformation method using Least Absolute Deviation,and the warp functions just consist affine transformation.Before warp generation, the user selects some set of handles, which may take the form of mesh nodes, segments, curves, points and so on. Each handle specifies an image feature, users control the deformation by manipulating the handles. As the user modifies the position and orientation of these handles, the image should deform in an intuitive fashion.The solution of Least Absolute Deviation has used enumeration method,which is more simplicity and exactly than other iteration method. When the character point of Image deformation is fewer,it will have good performances.