A New Sub-pixel Edge Detection Method of Color Images

Modern image measurement needs to take full advantage of sub-pixel edge information of images. This paper presents a sub-pixel edge detection method of color images based on image dimensionality reduction and the least square method. First of all we get the pixel level edge by Ostu algorithm and then combine gray processing algorithm based on color spatial distance and the least square method for sub-pixel edge location. Experimental results show that the algorithm positioning accuracy can reach 0.13 pixel which provides a basis for the selection of color image sub-pixel edge positioning. DOI : http://dx.doi.org/10.11591/telkomnika.v12i5.492