matlab灰度图上色代码改进

原例程代码地址

例程代码

代码简要分析

1.首先将灰度图转换为特殊的rgb图像（将灰度单通道扩展为rgb三通道）
2.将rgb三通道转换为Ycrbr三通道
3.根据Y分量对两张图进行匹配，用彩图Y分量最接近的点来给灰度图上的点上色

代码改进

function R=gray2rgb(img1,img2)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%　This Program converts a gray image ro RGB image based on the colors of the destination image. The　better the destination image match with the source gray image, the better the coloring will be. The program takes some time  as the searching time is high. You can decrease the searching time by taking　only samples from the used color image but quality may decrease. U can use jittered sampling for improving running speed. % 
%  You can use also use the attahed test images, Use the following combinations for better result　nature1.jpg(as img1) and nature2.jpg(as img2) or test1.jpg(as img1) and test2.jpg (as img2)　%
%　Usage: gray2rgb('nature1.jpg','nature2.jpg');  %

%  Authors : Jeny　Rajan , Chandrashekar　P.S %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% img1 - Source Image  (gray image)   
% img2 - Selected color image for coloring the gray image. 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Updates:
% 通过运用数据结构的知识减少遍历的次数，减少遍历的时间
%Author ：XiaoTian zou zi
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
tic
 clc;
 warning off;
 imt=imread(img1);
 ims=imread(img2);
 [sx sy sz]=size(imt);
 [tx ty tz]=size(ims);
 if sz~=1
     imt=rgb2gray(imt);
     imt=histeq(imt,256);          %均衡化后可减少椒盐噪声
 end
 if tz~=3
     disp ('img2 must be a color image (not indexed)');
 else
     imt(:,:,2)=imt(:,:,1);
     imt(:,:,3)=imt(:,:,1);

 % Converting to ycbcr color space
     nspace1=rgb2ycbcr(ims);
     nspace2= rgb2ycbcr(imt);
     mark=0;m=1;
     nspace3(m)=nspace1(1,1);
     disp('analysing......');
  for i=1:tx
      for j=1:ty
          mark=0;
          for k=1:m
              if(nspace1(i,j)==nspace3(k))
                  mark=1;
              end
          end                                     %delete the repeating color
          if(mark==0)
              nspace3(m+1,1)=nspace1(i,j,1);
              nspace3(m+1,2)=nspace1(i,j,2);
              nspace3(m+1,3)=nspace1(i,j,3);
              m=m+1;
          end  
      end
  end
     
     ms=double(nspace3(:,1));
     mt=double(nspace2(:,:,1));
     m1=max(ms);
     m2=min(ms);
     m3=max(max(mt));
     m4=min(min(mt));
     d1=m1-m2;
     d2=m3-m4;
 % Normalization
     dx1=ms;
     dx2=mt;
     dx1=(dx1*255)/(255-d1);
     dx2=(dx2*255)/(255-d2);
     
     
 %Luminance Comparison
     disp('Please wait..................');
     for i=1:sx
         for j=1:sy
              iy=dx2(i,j);
              tmp=abs(dx1-iy);
              ck=min(tmp);
              for r=1:m
                  
                      if(tmp(r)==ck)
                          break
                      end
                 
              end
                  nimage(i,j,2)=nspace3(r,2);
                  nimage(i,j,3)=nspace3(r,3);
                  nimage(i,j,1)=nspace2(i,j,1);           
            
          end
      end
     rslt=ycbcr2rgb(nimage);
     figure,imshow(uint8(imt));title('被上色图');
     figure,imshow(uint8(rslt));title('上色后的图');
     R=uint8(rslt);
     toc
 end  

原例程代码在做匹配时，每一个灰度图的点都要遍历一次彩色图大小的矩阵，看到例程的评论经常提到跑一遍程序需要花很长的时间。

因此考虑在源代码的基础上做一些修改，思路如下：

1.在匹配前先对彩色图像矩阵进行处理，将彩色图的各个像素点进行归类处理，得到一个只保存了彩图有差异的点矩阵。
2.用提取后的矩阵来进行像素匹配。
3.从程序运行角度来看，虽然彩图简化会使程序多一个遍历全图的时间，但是减少了彩图矩阵的大小，假设简化后的像素点减少了n个，而灰度图的像素点为x，那么总共就会减少n×x个像素比较的时间。

由于这是当年大二暑假编写的代码，最近看着还有很多和我一样的同学被运行时间过慢的问题而困扰，于是决定凭着回忆来记叙一下当时解决办法，可能细节上面有所疏漏，希望能够帮助到大家。