IMAGE TRANSFORMATION METHOD FOR DETERMINING KERNEL MOTION POSITIONS IN 3 DIMENSIONS

An image transformation method for determining kernel coordinates in t hree dimensions was developed. Instead of using two cameras to recover the lost dimension caused by projecting a three-dimensional object on to a two-dimensional film plane, one camera (with the help of a mirror and strobe light) was used to take photos of discrete kernel motion t rajectory with two views - one real image and one virtual image from t he mirror. The image transformations were derived to determine kernel motion positions in three dimensions on the basis of the real kernel i mages and virtual kernel images from the photos.