Provided is an image processing apparatus. In the case of a multi-view image, the image processing apparatus may warp a color image and/or a depth image of a reference view to correspond to a target view. An occlusion region may be extracted by comparing the warped color image and/or depth image of the reference view with a color image and/or a depth image of the target view. The image processing apparatus may encode the extracted occlusion region information without loss, and decrease an amount of data in the multi-view image.