In automated apple sorting and grading applications, one persistent problem is to identify apple stem-ends and calyxes from defects. To solve this problem, a Three-dimensional (3D) Shape Enhanced Transform (SET) approach is presented. The SET method enhances the apple stem-end/calyx area according to the 3D surface gradient difference between the stem-end/calyx and the apple surface region. In addition, the proposed SET approach does not depend on the location of the stem-end/calyx on the apple surface, making it more suitable for apples orientated randomly. SET is also an automated and robust method, which detects the apple stem-end/calyx without any human intervention, and performs well with noise and even incomplete image data. A total of 232 Golden Delicious apple images were tested, and an overall detection rate of 93.97% was achieved.