Orientation Detection Using the Features of Vectors Constructed by the Edge Points and Gravity Centre of an Image

In this paper, an algorithm for detecting orientation of an object from a two-dimensional binary image which has an uniform density is presented. A curve derived from vectors based on the gravity centre and the edge points of an object’s image is generated from this algorithm to represent the invariant feature of rotation of an object. The results show that the orientation of an examined object with reference to its initial position can be found. A related technique to detect edge is also studied in the paper.