An image processing apparatus includes: a modeling unit, in video image data composed of frame images having an X direction and a Y direction perpendicular to the X direction, separating a motion from a standard image, which is one frame before a reference image subjected to a process, until the reference image into motions in the X and Y directions and representing the motions in the X and Y directions respectively as vector fields composed of a translation component composed of identical magnitude and direction regardless of positions in the X and Y directions, an X dependent component primarily changing with the position in the X direction, and a Y dependent component primarily changing with the position in the Y direction; and a calculation unit respectively calculating the translation component and the X and Y dependent components corresponding to the X and Y directions represented by the modeling unit.