图像处理（三）——中国传统工艺画风格的图像转换应用
中国传统工艺画风格的图像转换应用
1. 项目介绍
希望做一个应用，输入为普通RGB或HSI图像，输出为带有中国特色的蛋壳粘贴画。使用的技术有Matlab、superpixel算法等。
蛋壳粘贴画示意
如图，蛋壳粘贴画每一块破碎的蛋壳就像是一个超像素。项目构思为：首先利用superpixel算法找出超像素，对原图像进行分割；在每个超像素内求各像素点的均值，作为整块超像素的颜色值。这样每个超像素作为一块破碎的蛋壳，一幅蛋壳粘贴画就成功生成了！
2. 实现过程
在网上找到了superpixel算法的C语言实现，同时给出了调用C（语言实现的）函数的.m文件。但是没有找到Matlab代码实现的superpixel算法。于是想跟找到的代码一样，利用Matlab运行 .c文件。之后对 .c文件进行阅读、修改、增加自己的功能，以达到项目的目标。
2.1 尝试用Matlab运行 .c文件
我的Matlab要想在.m文件中调用C（语言实现的）函数还需要进行配置。
网上的教程说需要使用mex命令，以某种方式编译过C文件后才可以被Matlab运行。尝试运行mex命令：
错误的原因是Windows SDK 7.1已经安装，但是它所携带的编译器没有安装。提示安装.NET Framework 4.0，之后重新运行Windows SDK 7.1的安装程序，勾选“Visual C++ Compilers”，这样Windows SDK 7.1的安装程序就会安装编译器。
另一种方法是“Install MinGW-w64 Compiler”，已经试过安装，但没有成功。
后来我发现其他人的电脑直接就能mex -setup成功，不会报错，因为装了VS2013；而我的电脑装的是VS2015，所以版本新并不一定好...
2.2 安装.NET Framework 4.0
从http://www.microsoft.com/zh-cn/download/details.aspx?id=17718下载.NET Framework 4.0独立安装包，进行安装。出现以下提示。
意思是.NET Framework 4.0已经安装好了。我的系统是win10专业版。
2.3 再次运行Windows SDK 7.1的安装程序
安装程序是从http://www.microsoft.com/en-us/download/details.aspx?id=8279下载的。运行结果：
“Visual C++ Compilers”选项不可勾选，将鼠标移到这个选项上显示“This feature is disabled because required .NET Framework 4.0 is not installed”。错误原因又是.NET Framework 4.0没有安装。刚才明明提示.NET Framework 4.0已经是系统的一部分。我的系统是win10专业版。
不知如何是好。
问题应该在于我的电脑里装了.NET Framework 4.6，它是.NET Framework 4.0的升级版，无法在有.NET 4.6的情况下安装.NET 4.0，因为他们本质上就是一个东西！
问题困扰了我那么久的原因：
a. 电脑上安装的是VS2015，而不是VS2013；
b. 原则上.NET 4.6应该兼容.NET 4.0，但是“Visual C++ Compilers”只认.NET 4.0；
c. 要先卸载.NET 4.6才能安装.NET 4.0，但是系统中.NET 4.6卸载不干净！
2.4 用Matlab运行简单的 .c文件
最后终于可以在我的matlab中运行 .c文件了，解决方法是直接安装“VC-Compiler-KB2519277.exe”（下载地址）。之后在matlab命令行中先后执行以下两条命令:
mex -setup
mex -setup C++
编译器即配置完成！
测试一下，编写一个简单的add.c，代码如下：
#include "mex.h" // 使用MEX文件必须包含的头文件 // 执行具体工作的C函数 double add(double x, double y) { return x + y; } // MEX文件接口函数 void mexFunction(int nlhs,mxArray *plhs[], int nrhs,const mxArray *prhs[]) { double *a; double b, c; plhs[0] = mxCreateDoubleMatrix(1, 1, mxREAL); a = mxGetPr(plhs[0]); b = *(mxGetPr(prhs[0])); c = *(mxGetPr(prhs[1])); *a = add(b, c); }
在matlab命令行中执行命令:
mex add.c
之后就可以在matlab中直接调用add函数了，如下图：
以上add.c 并不是单纯的C语言代码，内部代码多出了一个MEX文件接口函数。
2.4 用Matlab运行自己的 .c文件
这是从网站上下载的slicmex.c，原文件下载，这里的slicmex.c经过了修改。
slicmex.c:
1 #include <mex.h> 2 #include <stdio.h> 3 #include <math.h> 4 #include <float.h> 5 6 void rgbtolab(int* rin, int* gin, int* bin, int sz, double* lvec, double* avec, double* bvec) 7 { 8 int i; int sR, sG, sB; 9 double R,G,B; 10 double X,Y,Z; 11 double r, g, b; 12 const double epsilon = 0.008856; //actual CIE standard 13 const double kappa = 903.3; //actual CIE standard 14 15 const double Xr = 0.950456; //reference white 16 const double Yr = 1.0; //reference white 17 const double Zr = 1.088754; //reference white 18 double xr,yr,zr; 19 double fx, fy, fz; 20 double lval,aval,bval; 21 22 for(i = 0; i < sz; i++) 23 { 24 sR = rin[i]; sG = gin[i]; sB = bin[i]; 25 R = sR/255.0; 26 G = sG/255.0; 27 B = sB/255.0; 28 29 if(R <= 0.04045) r = R/12.92; 30 else r = pow((R+0.055)/1.055,2.4); 31 if(G <= 0.04045) g = G/12.92; 32 else g = pow((G+0.055)/1.055,2.4); 33 if(B <= 0.04045) b = B/12.92; 34 else b = pow((B+0.055)/1.055,2.4); 35 36 X = r*0.4124564 + g*0.3575761 + b*0.1804375; 37 Y = r*0.2126729 + g*0.7151522 + b*0.0721750; 38 Z = r*0.0193339 + g*0.1191920 + b*0.9503041; 39 40 //------------------------ 41 // XYZ to LAB conversion 42 //------------------------ 43 xr = X/Xr; 44 yr = Y/Yr; 45 zr = Z/Zr; 46 47 if(xr > epsilon) fx = pow(xr, 1.0/3.0); 48 else fx = (kappa*xr + 16.0)/116.0; 49 if(yr > epsilon) fy = pow(yr, 1.0/3.0); 50 else fy = (kappa*yr + 16.0)/116.0; 51 if(zr > epsilon) fz = pow(zr, 1.0/3.0); 52 else fz = (kappa*zr + 16.0)/116.0; 53 54 lval = 116.0*fy-16.0; 55 aval = 500.0*(fx-fy); 56 bval = 200.0*(fy-fz); 57 58 lvec[i] = lval; avec[i] = aval; bvec[i] = bval; 59 } 60 } 61 62 void getLABXYSeeds(int STEP, int width, int height, int* seedIndices, int* numseeds) 63 { 64 const bool hexgrid = false; 65 int n; 66 int xstrips, ystrips; 67 int xerr, yerr; 68 double xerrperstrip,yerrperstrip; 69 int xoff,yoff; 70 int x,y; 71 int xe,ye; 72 int seedx,seedy; 73 int i; 74 75 xstrips = (0.5+(double)(width)/(double)(STEP)); 76 ystrips = (0.5+(double)(height)/(double)(STEP)); 77 78 xerr = width - STEP*xstrips;if(xerr < 0){xstrips--;xerr = width - STEP*xstrips;} 79 yerr = height - STEP*ystrips;if(yerr < 0){ystrips--;yerr = height- STEP*ystrips;} 80 81 xerrperstrip = (double)(xerr)/(double)(xstrips); 82 yerrperstrip = (double)(yerr)/(double)(ystrips); 83 84 xoff = STEP/2; 85 yoff = STEP/2; 86 87 n = 0; 88 for( y = 0; y < ystrips; y++ ) 89 { 90 ye = y*yerrperstrip; 91 for( x = 0; x < xstrips; x++ ) 92 { 93 xe = x*xerrperstrip; 94 seedx = (x*STEP+xoff+xe); 95 if(hexgrid){ seedx = x*STEP+(xoff<<(y&0x1))+xe; if(seedx >= width)seedx = width-1; }//for hex grid sampling 96 seedy = (y*STEP+yoff+ye); 97 i = seedy*width + seedx; 98 seedIndices[n] = i; 99 n++; 100 } 101 } 102 *numseeds = n; 103 } 104 105 void PerformSuperpixelSLIC(double* lvec, double* avec, double* bvec, double* kseedsl, double* kseedsa, double* kseedsb, double* kseedsx, double* kseedsy, int width, int height, int numseeds, int* klabels, int STEP, double compactness) 106 { 107 int x1, y1, x2, y2; 108 double l, a, b; 109 double dist; 110 double distxy; 111 int itr; 112 int n; 113 int x,y; 114 int i; 115 int ind; 116 int r,c; 117 int k; 118 int sz = width*height; 119 const int numk = numseeds; 120 int offset = STEP; 121 122 double* clustersize = mxMalloc(sizeof(double)*numk); 123 double* inv = mxMalloc(sizeof(double)*numk); 124 double* sigmal = mxMalloc(sizeof(double)*numk); 125 double* sigmaa = mxMalloc(sizeof(double)*numk); 126 double* sigmab = mxMalloc(sizeof(double)*numk); 127 double* sigmax = mxMalloc(sizeof(double)*numk); 128 double* sigmay = mxMalloc(sizeof(double)*numk); 129 double* distvec = mxMalloc(sizeof(double)*sz); 130 double invwt = 1.0/((STEP/compactness)*(STEP/compactness)); 131 132 for( itr = 0; itr < 10; itr++ ) 133 { 134 for(i = 0; i < sz; i++){distvec[i] = DBL_MAX;} 135 136 for( n = 0; n < numk; n++ ) 137 { 138 x1 = kseedsx[n]-offset; if(x1 < 0) x1 = 0; 139 y1 = kseedsy[n]-offset; if(y1 < 0) y1 = 0; 140 x2 = kseedsx[n]+offset; if(x2 > width) x2 = width; 141 y2 = kseedsy[n]+offset; if(y2 > height) y2 = height; 142 143 for( y = y1; y < y2; y++ ) 144 { 145 for( x = x1; x < x2; x++ ) 146 { 147 i = y*width + x; 148 149 l = lvec[i]; 150 a = avec[i]; 151 b = bvec[i]; 152 153 dist = (l - kseedsl[n])*(l - kseedsl[n]) + 154 (a - kseedsa[n])*(a - kseedsa[n]) + 155 (b - kseedsb[n])*(b - kseedsb[n]); 156 157 distxy = (x - kseedsx[n])*(x - kseedsx[n]) + (y - kseedsy[n])*(y - kseedsy[n]); 158 159 dist += distxy*invwt; 160 161 if(dist < distvec[i]) 162 { 163 distvec[i] = dist; 164 klabels[i] = n; 165 } 166 } 167 } 168 } 169 //----------------------------------------------------------------- 170 // Recalculate the centroid and store in the seed values 171 //----------------------------------------------------------------- 172 for(k = 0; k < numk; k++) 173 { 174 sigmal[k] = 0; 175 sigmaa[k] = 0; 176 sigmab[k] = 0; 177 sigmax[k] = 0; 178 sigmay[k] = 0; 179 clustersize[k] = 0; 180 } 181 182 ind = 0; 183 for( r = 0; r < height; r++ ) 184 { 185 for( c = 0; c < width; c++ ) 186 { 187 if(klabels[ind] >= 0) 188 { 189 sigmal[klabels[ind]] += lvec[ind]; 190 sigmaa[klabels[ind]] += avec[ind]; 191 sigmab[klabels[ind]] += bvec[ind]; 192 sigmax[klabels[ind]] += c; 193 sigmay[klabels[ind]] += r; 194 clustersize[klabels[ind]] += 1.0; 195 } 196 ind++; 197 } 198 } 199 200 {for( k = 0; k < numk; k++ ) 201 { 202 if( clustersize[k] <= 0 ) clustersize[k] = 1; 203 inv[k] = 1.0/clustersize[k];//computing inverse now to multiply, than divide later 204 }} 205 206 {for( k = 0; k < numk; k++ ) 207 { 208 kseedsl[k] = sigmal[k]*inv[k]; 209 kseedsa[k] = sigmaa[k]*inv[k]; 210 kseedsb[k] = sigmab[k]*inv[k]; 211 kseedsx[k] = sigmax[k]*inv[k]; 212 kseedsy[k] = sigmay[k]*inv[k]; 213 }} 214 } 215 mxFree(sigmal); 216 mxFree(sigmaa); 217 mxFree(sigmab); 218 mxFree(sigmax); 219 mxFree(sigmay); 220 mxFree(clustersize); 221 mxFree(inv); 222 mxFree(distvec); 223 } 224 225 void EnforceSuperpixelConnectivity(int* labels, int width, int height, int numSuperpixels,int* nlabels, int* finalNumberOfLabels) 226 { 227 int i,j,k; 228 int n,c,count; 229 int x,y; 230 int ind; 231 int oindex, adjlabel; 232 int label; 233 const int dx4[4] = {-1, 0, 1, 0}; 234 const int dy4[4] = { 0, -1, 0, 1}; 235 const int sz = width*height; 236 const int SUPSZ = sz/numSuperpixels; 237 int* xvec = mxMalloc(sizeof(int)*SUPSZ*10); 238 int* yvec = mxMalloc(sizeof(int)*SUPSZ*10); 239 240 for( i = 0; i < sz; i++ ) nlabels[i] = -1; 241 oindex = 0; 242 adjlabel = 0;//adjacent label 243 label = 0; 244 for( j = 0; j < height; j++ ) 245 { 246 for( k = 0; k < width; k++ ) 247 { 248 if( 0 > nlabels[oindex] ) 249 { 250 nlabels[oindex] = label; 251 //-------------------- 252 // Start a new segment 253 //-------------------- 254 xvec[0] = k; 255 yvec[0] = j; 256 //------------------------------------------------------- 257 // Quickly find an adjacent label for use later if needed 258 //------------------------------------------------------- 259 {for( n = 0; n < 4; n++ ) 260 { 261 int x = xvec[0] + dx4[n]; 262 int y = yvec[0] + dy4[n]; 263 if( (x >= 0 && x < width) && (y >= 0 && y < height) ) 264 { 265 int nindex = y*width + x; 266 if(nlabels[nindex] >= 0) adjlabel = nlabels[nindex]; 267 } 268 }} 269 270 count = 1; 271 for( c = 0; c < count; c++ ) 272 { 273 for( n = 0; n < 4; n++ ) 274 { 275 x = xvec[c] + dx4[n]; 276 y = yvec[c] + dy4[n]; 277 278 if( (x >= 0 && x < width) && (y >= 0 && y < height) ) 279 { 280 int nindex = y*width + x; 281 282 if( 0 > nlabels[nindex] && labels[oindex] == labels[nindex] ) 283 { 284 xvec[count] = x; 285 yvec[count] = y; 286 nlabels[nindex] = label; 287 count++; 288 } 289 } 290 291 } 292 } 293 //------------------------------------------------------- 294 // If segment size is less then a limit, assign an 295 // adjacent label found before, and decrement label count. 296 //------------------------------------------------------- 297 if(count <= SUPSZ >> 2) 298 { 299 for( c = 0; c < count; c++ ) 300 { 301 ind = yvec[c]*width+xvec[c]; 302 nlabels[ind] = adjlabel; 303 } 304 label--; 305 } 306 label++; 307 } 308 oindex++; 309 } 310 } 311 *finalNumberOfLabels = label; 312 313 mxFree(xvec); 314 mxFree(yvec); 315 } 316 317 void mexFunction(int nlhs, mxArray *plhs[], 318 int nrhs, const mxArray *prhs[]) 319 { 320 int width; 321 int height; 322 int sz; 323 int i, ii; 324 int x, y; 325 int* rin; int* gin; int* bin; 326 int* klabels; 327 int* clabels; 328 double* lvec; double* avec; double* bvec; 329 int step; 330 int* seedIndices; 331 int numseeds; 332 double* kseedsx;double* kseedsy; 333 double* kseedsl;double* kseedsa;double* kseedsb; 334 int k; 335 const mwSize* dims;//int* dims; 336 int* outputNumSuperpixels; 337 int* outlabels; 338 int finalNumberOfLabels; 339 unsigned char* imgbytes; 340 int numelements; 341 int numSuperpixels = 200;//default value 342 double compactness = 10;//default value 343 int numdims; 344 345 if (nrhs < 1) { 346 mexErrMsgTxt("At least one argument is required.") ; 347 } else if(nrhs > 3) { 348 mexErrMsgTxt("Too many input arguments."); 349 } 350 if(nlhs!=2) { 351 mexErrMsgIdAndTxt("SLIC:nlhs","Two outputs required, a labels and the number of labels, i.e superpixels."); 352 } 353 //--------------------------- 354 numelements = mxGetNumberOfElements(prhs[0]) ; 355 numdims = mxGetNumberOfDimensions(prhs[0]) ; 356 dims = mxGetDimensions(prhs[0]) ; 357 imgbytes = (unsigned char*)mxGetData(prhs[0]) ;//mxGetData returns a void pointer, so cast it 358 width = dims[1]; height = dims[0];//Note: first dimension provided is height and second is width 359 sz = width*height; 360 //--------------------------- 361 numSuperpixels = mxGetScalar(prhs[1]); 362 compactness = mxGetScalar(prhs[2]); 363 364 //--------------------------- 365 // Allocate memory 366 //--------------------------- 367 rin = mxMalloc( sizeof(int) * sz ) ; 368 gin = mxMalloc( sizeof(int) * sz ) ; 369 bin = mxMalloc( sizeof(int) * sz ) ; 370 lvec = mxMalloc( sizeof(double) * sz ) ; 371 avec = mxMalloc( sizeof(double) * sz ) ; 372 bvec = mxMalloc( sizeof(double) * sz ) ; 373 klabels = mxMalloc( sizeof(int) * sz );//original k-means labels 374 clabels = mxMalloc( sizeof(int) * sz );//corrected labels after enforcing connectivity 375 seedIndices = mxMalloc( sizeof(int) * sz ); 376 377 //--------------------------- 378 // Perform color conversion 379 //--------------------------- 380 //if(2 == numdims) 381 if(numelements/sz == 1)//if it is a grayscale image, copy the values directly into the lab vectors 382 { 383 for(x = 0, ii = 0; x < width; x++)//reading data from column-major MATLAB matrics to row-major C matrices (i.e perform transpose) 384 { 385 for(y = 0; y < height; y++) 386 { 387 i = y*width+x; 388 lvec[i] = imgbytes[ii]; 389 avec[i] = imgbytes[ii]; 390 bvec[i] = imgbytes[ii]; 391 ii++; 392 } 393 } 394 } 395 else//else covert from rgb to lab 396 { 397 for(x = 0, ii = 0; x < width; x++)//reading data from column-major MATLAB matrics to row-major C matrices (i.e perform transpose) 398 { 399 for(y = 0; y < height; y++) 400 { 401 i = y*width+x; 402 rin[i] = imgbytes[ii]; 403 gin[i] = imgbytes[ii+sz]; 404 bin[i] = imgbytes[ii+sz+sz]; 405 ii++; 406 } 407 } 408 rgbtolab(rin,gin,bin,sz,lvec,avec,bvec); 409 } 410 //--------------------------- 411 // Find seeds 412 //--------------------------- 413 step = sqrt((double)(sz)/(double)(numSuperpixels))+0.5; 414 getLABXYSeeds(step,width,height,seedIndices,&numseeds); 415 416 kseedsx = mxMalloc( sizeof(double) * numseeds ) ; 417 kseedsy = mxMalloc( sizeof(double) * numseeds ) ; 418 kseedsl = mxMalloc( sizeof(double) * numseeds ) ; 419 kseedsa = mxMalloc( sizeof(double) * numseeds ) ; 420 kseedsb = mxMalloc( sizeof(double) * numseeds ) ; 421 for(k = 0; k < numseeds; k++) 422 { 423 kseedsx[k] = seedIndices[k]%width; 424 kseedsy[k] = seedIndices[k]/width; 425 kseedsl[k] = lvec[seedIndices[k]]; 426 kseedsa[k] = avec[seedIndices[k]]; 427 kseedsb[k] = bvec[seedIndices[k]]; 428 } 429 //--------------------------- 430 // Compute superpixels 431 //--------------------------- 432 PerformSuperpixelSLIC(lvec, avec, bvec, kseedsl,kseedsa,kseedsb,kseedsx,kseedsy,width,height,numseeds,klabels,step,compactness); 433 //--------------------------- 434 // Enforce connectivity 435 //--------------------------- 436 EnforceSuperpixelConnectivity(klabels,width,height,numSuperpixels,clabels,&finalNumberOfLabels); 437 //--------------------------- 438 // Assign output labels 439 //--------------------------- 440 plhs[0] = mxCreateNumericMatrix(height,width,mxINT32_CLASS,mxREAL); 441 outlabels = mxGetData(plhs[0]); 442 for(x = 0, ii = 0; x < width; x++)//copying data from row-major C matrix to column-major MATLAB matrix (i.e. perform transpose) 443 { 444 for(y = 0; y < height; y++) 445 { 446 i = y*width+x; 447 outlabels[ii] = clabels[i]; 448 ii++; 449 } 450 } 451 //--------------------------- 452 // Assign number of labels/seeds 453 //--------------------------- 454 plhs[1] = mxCreateNumericMatrix(1,1,mxINT32_CLASS,mxREAL); 455 outputNumSuperpixels = (int*)mxGetData(plhs[1]);//gives a void*, cast it to int* 456 *outputNumSuperpixels = finalNumberOfLabels; 457 //--------------------------- 458 // Deallocate memory 459 //--------------------------- 460 mxFree(rin); 461 mxFree(gin); 462 mxFree(bin); 463 mxFree(lvec); 464 mxFree(avec); 465 mxFree(bvec); 466 mxFree(klabels); 467 mxFree(clabels); 468 mxFree(seedIndices); 469 mxFree(kseedsx); 470 mxFree(kseedsy); 471 mxFree(kseedsl); 472 mxFree(kseedsa); 473 mxFree(kseedsb); 474 }
在matlab命令行中执行命令:
mex slicmex.c
就可以直接调用slicmex()函数了。
2.5 编写matlab代码
myslic_function2.m代码:
%====================================================================== % 此函数将图像分为超像素块，块儿内颜色取值一样，为之前块儿内颜色的均值 %====================================================================== function imgout=myslic_function2(imgin, number_of_sp, compactness_factor) imgout = imgin; [m,n] = size(imgin); n=n/3; [labels, numlabels] = slicmex(imgin,number_of_sp,compactness_factor);%numlabels is the same as number of superpixels imgin = int32(imgin); labelstmp = labels; % 以下所有代码是将RGB图像的三个通道，以labels为模板，分别求每个超像素内颜色的均值。 img = imgin(:,:,1); img1 = img; for i = 1 : m for j = 1 : n if (labelstmp(i, j) >= 0) current = labelstmp(i, j); sums = 0; count = 0; %average= 0; % 找到一个块，计算块儿内颜色的均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) sums = sums + img(ii, jj); count = count + 1; end end end average = sums / count; % 将块儿内颜色设为均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) img1(ii, jj) = average; labelstmp(ii, jj) = -1; end end end end end end imgout(:,:,1) = img1; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% labelstmp = labels; img = imgin(:,:,2); img1 = img; for i = 1 : m for j = 1 : n if (labelstmp(i, j) >= 0) current = labelstmp(i, j); sums = 0; count = 0; %average= 0; % 找到一个块，计算块儿内颜色的均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) sums = sums + img(ii, jj); count = count + 1; end end end average = sums / count; % 将块儿内颜色设为均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) img1(ii, jj) = average; labelstmp(ii, jj) = -1; end end end end end end imgout(:,:,2) = img1; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% labelstmp = labels; img = imgin(:,:,3); img1 = img; for i = 1 : m for j = 1 : n if (labelstmp(i, j) >= 0) current = labelstmp(i, j); sums = 0; count = 0; %average= 0; % 找到一个块，计算块儿内颜色的均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) sums = sums + img(ii, jj); count = count + 1; end end end average = sums / count; % 将块儿内颜色设为均值 for ii = 1 : m for jj = 1 : n if (labelstmp(ii, jj) == current) img1(ii, jj) = average; labelstmp(ii, jj) = -1; end end end end end end imgout(:,:,3) = img1;
myslic_function2函数的效果为：
上面原图为img，myslic_function2函数处理过的图为img2，其关系为
img2 = myslic_function2(img, 1000, 100);
myslic_function2函数接受3个参数，分别是原图、结果想要的超像素的个数、Compactness factor（值越大超像素块儿边界越光滑）。
要想获得蛋壳粘贴画的效果，还应在超像素块儿间加上裂纹，当然裂纹也通过对slic算法的结果进行再处理产生。
产生裂纹的代码段：
img = imread('06.jpg'); edgeline = rgb2gray(img); [labels, numlabels] = slicmex(img, 1000, 100); [m,n]=size(labels); for ii = 2 : m-1 for jj = 2 : n-1 if (labels(ii, jj) == labels(ii-1, jj)&&labels(ii, jj) == labels(ii+1, jj)&&labels(ii, jj) == labels(ii, jj-1)&&labels(ii, jj) == labels(ii, jj+1) ) edgeline(ii, jj)=0; else edgeline(ii, jj)=255; end end end % 对已经产生的裂纹进行高斯滤波，之后再进行阈值过滤 sigma = 1.3; gausFilter = fspecial('gaussian',[5 5],sigma); edgeline=imfilter(edgeline,gausFilter,'replicate'); edgeline(edgeline>100) = 255; edgeline(edgeline<=100) = 0; figure(3); imshow(edgeline);
产生的裂纹效果：
最后，将产生裂纹加在经过myslic_function2函数处理后的图片上，代码如下：
img2(:,:,1)=img2(:,:,1) - edgeline; img2(:,:,2)=img2(:,:,2) - edgeline; img2(:,:,3)=img2(:,:,3) - edgeline;
最后的结果：
挺萌的还。附项目目录下载。其中image_shade1.m是运行的图形界面。
