STUDY ON IFD CALCULATED THEORY AND NATURE OF TRUE COLOR IMAGE

Abstract

In this paper, a new calculated method of image fractal dimension (IFD) of true color image has been presented in order to solve the problem that laser scanning profilimeter, applied to measuring fractal characteristic of rough surface, has too small scanning field to obtain three-dimensional coordinates of larger area surface. The method can also characterize indirectly surface profile by quantifying roughness of true color surface built in terms of connecting endpoint of the vector with three components of red, green and blue. The correctness and feasibility of the IFD method have been verified by scanning test on rough surface using larger-scale laser profilimeter. Comparison between conventional dimension theory presented in previous study and IFD proposed in the paper has shown that IFD method has the advantage of simplification, convenience and large scanning capacity, and is progressing in the field of surface configuration investigation. Furthermore, fractal characteristic of binary image, grey-scale image as well as true color image is researched to probe their physical nature. Comparison among IFD of three kinds of image implies that, IFD of true color image is largest, IFD of grey-scale image is second largest, and IFD of binary image varying within the range of 1–2 is the smallest. From the phenomena, it is indicated that: (i) IFD of binary image is dimension of curved line; (ii) for IFD of grey-scale image, only effect of luminance is considered; (iii) for IFD of true color image, contribution of hue, saturation and luminance is sufficiently considered, and therefore it is just "complete" IFD.