武汉理工大学数字图像处理英文课件 Chapter3（2）.pptVIP

Co-Histogram Matlab function cohist Example of cohist Co-Histogram Plots Color Component Co-Histograms * A co-histogram is the joint probability distribution of pixel values in two separate images. For every pixel value a in image g1, generate a column histogram of pixel values in image g2. Equivalently, for every pixel value b in image g2, generate a row histogram of pixel values in image g1. h(a,b) counts the frequency of occurrence of pixel value a in image g1 and pixel value b in image g2. To provide significant counts, the bins are usually coarser than for individual image histograms. function [nout] = cohist(x,y,m,d) % Co-occurance matrix or joint PDF. % N = COHIST(X,Y) correlates the elements of X and Y into m x m equally % spaced containers and returns the number of elements in each container. % X and Y are matrices of type double with range [0,1] with the same % number of rows and columns. % % N = COHIST(Y,X,M), where M is a scalar, uses M bins. % % N = COHIST(Y,X,M,D), where D is a scalar, uses D decades of scaling. % % COHIST(...) without output arguments produces a histogram image of % the results. % R.M. Clodfelter, June 1998 % Copyright (c) 1998 by R.M. Clodfelter % $Revision: 1.00 $ $Date: 1998/06/16 $ gray = imread(kitchen2.tif); rgb = imread(kitchen1.tif); gray = im2double(gray); rgb = im2double(rgb); g1 = gray; g2 = rgb2gray(rgb); cohist(g1,g2); r = corr2(g1,g2) r = 0.9993 [X,Y]=meshgrid(linspace(0,255,33)); axis([0 255 0 255 0 8000]); mesh(X,Y,nn); Projections are histograms of g1 and g2. cohist(g1,g2); red blue red green green blue *