from: http://blog.csdn.net/jia20003/article/details/7757276
图像处理之应用篇-大米计数续
背景介绍:
请看博客文章《图像处理之简单综合实例(大米计数)》
其实拍出来的照片更多的是含有大米颗粒相互接触,甚至于有点重叠的照片
要准确计算大米的颗粒数非常困难,通过图像形态学开闭操作,腐蚀等手
段尝试以后效果不是很好。最终发现一种简单明了但是有微小误差的计数
方法。照相机图片:
算法思想:
主要是利用连通区域发现算法,发现所有连通区域,使用二分法,截取较小
部分的连通区域集合,求取平均连通区域面积,根据此平均连通区域面积,
作为单个大米大小,从而求取出粘连部分的大米颗粒数,完成对整个大米
数目的统计:
缺点:
平均连通区域面积的计算受制于两个因素,一个是最小连通区域集合的选取算法,
二个样本数量。算法结果跟实际结果有一定的误差,但是误差在1%左右。
程序算法代码详解
将输入图像转换为黑白二值图像,求得连通区域的算法代码如下:
src = super.filter(src,null);
getRGB(src, 0, 0, width,height, inPixels );
FastConnectedComponentLabelAlgfccAlg = new FastConnectedComponentLabelAlg();
fccAlg.setBgColor(0);
int[] outData = fccAlg.doLabel(inPixels, width, height);
获取平均大米颗粒连通区域的代码如下:
Integer[] values =labelMap.values().toArray(new Integer[0]);
Arrays.sort(values);
int minRiceNum = values.length/4;
float sum = 0;
for(int v=offset; v<minRiceNum +offset; v++) {
sum += values[v].intValue();
}
float minMeans = sum / (float)minRiceNum;
System.out.println(" minMeans = " + minMeans);
程序时序图如下:
程序运行效果如下:
实际大米颗粒数目为202,正确率为99%
完成大米数目统计的源代码如下(其它相关代码见以前的图像处理系列文章):
- public class FindRiceFilter extends BinaryFilter {
- private int sumRice;
- private int offset = 10;
- public int getSumRice() {
- return this.sumRice;
- }
- public void setOffset(int pos) {
- this.offset = pos;
- }
- @Override
- public BufferedImage filter(BufferedImage src, BufferedImage dest) {
- int width = src.getWidth();
- int height = src.getHeight();
- if ( dest == null )
- dest = createCompatibleDestImage( src, null );
- int[] inPixels = new int[width*height];
- int[] outPixels = new int[width*height];
- src = super.filter(src, null);
- getRGB(src, 0, 0, width, height, inPixels );
- FastConnectedComponentLabelAlg fccAlg = new FastConnectedComponentLabelAlg();
- fccAlg.setBgColor(0);
- int[] outData = fccAlg.doLabel(inPixels, width, height);
- // labels statistic
- HashMap<Integer, Integer> labelMap = new HashMap<Integer, Integer>();
- for(int d=0; d<outData.length; d++) {
- if(outData[d] != 0) {
- if(labelMap.containsKey(outData[d])) {
- Integer count = labelMap.get(outData[d]);
- count+=1;
- labelMap.put(outData[d], count);
- } else {
- labelMap.put(outData[d], 1);
- }
- }
- }
- Integer[] values = labelMap.values().toArray(new Integer[0]);
- Arrays.sort(values);
- int minRiceNum = values.length/4;
- float sum = 0;
- for(int v= offset; v<minRiceNum + offset; v++) {
- sum += values[v].intValue();
- }
- float minMeans = sum / (float)minRiceNum;
- System.out.println(" minMeans = " + minMeans);
- // try to find the max connected component
- Integer[] keys = labelMap.keySet().toArray(new Integer[0]);
- Arrays.sort(keys);
- int threshold = 10;
- ArrayList<Integer> listKeys = new ArrayList<Integer>();
- for(Integer key : keys) {
- if(labelMap.get(key) <=threshold){
- listKeys.add(key);
- } else {
- float xx = labelMap.get(key);
- float intPart = (float)Math.floor(xx / minMeans + 0.5f);
- sumRice += intPart;
- }
- }
- System.out.println( "Number of rice = " + sumRice);
- // sumRice = keys.length - listKeys.size();
- // calculate means of pixel
- int index = 0;
- for(int row=0; row<height; row++) {
- int ta = 0, tr = 0, tg = 0, tb = 0;
- for(int col=0; col<width; col++) {
- index = row * width + col;
- ta = (inPixels[index] >> 24) & 0xff;
- tr = (inPixels[index] >> 16) & 0xff;
- tg = (inPixels[index] >> 8) & 0xff;
- tb = inPixels[index] & 0xff;
- if(outData[index] != 0 && validRice(outData[index], listKeys)) {
- tr = tg = tb = 255;
- } else {
- tr = tg = tb = 0;
- }
- outPixels[index] = (ta << 24) | (tr << 16) | (tg << 8) | tb;
- }
- }
- setRGB( dest, 0, 0, width, height, outPixels );
- return dest;
- }
- private boolean validRice(int i, ArrayList<Integer> listKeys) {
- for(Integer key : listKeys) {
- if(key == i) {
- return false;
- }
- }
- return true;
- }
- }