An approach on characteristic lines extracting of gray images using moment of inertia

A novel method on characteristic lines extracting of digital images is presented in this paper by using the moment of inertia theory, which is used to extract the characteristic lines in digital images through treating them as rigid bodies. The characteristic lines process scale and rotation invariant inherently. Experimental results manifest that using the angle difference between characteristic lines in an original image and a rotated and scale changed image. The rotation angle of this image can be calculated correctly. The characteristic line can be also used to detect the symmetry axis of a symmetry region.