An image position matching apparatus and an image processing apparatus capable of accurately matching the positions of a three-dimensional past image and a two-dimensional present image, even for cases in which fluctuation in the position of the subject has occurred when the respective images were obtained. An image reducing means reduces to 1/10 size a three-dimensional past image data and a two-dimensional current image data that have been inputted thereto. The reduced three-dimensional past image data and the reduced two-dimensional current image data are inputted to a position matching image obtaining means from the reducing means. After the reduced three-dimensional image data has been subjected to a three-dimensional affine transform by the position matching image obtaining means, a two-dimensional past image data having a high correlation with the reduced two-dimensional present image is obtained utilizing a projection transform by perspective projection.