Android图片二值化算法

转自：http://gqdy365.iteye.com/blog/1559180

	public void binarization(Bitmap img) {
		width = img.getWidth();
		height = img.getHeight();
		int area = width * height;
		int gray[][] = new int[width][height];
		int average = 0;// 灰度平均值
		int graysum = 0;
		int graymean = 0;
		int grayfrontmean = 0;
		int graybackmean = 0;
		int pixelGray;
		int front = 0;
		int back = 0;
		int[] pix = new int[width * height];
		img.getPixels(pix, 0, width, 0, 0, width, height);
		for (int i = 1; i < width; i++) { // 不算边界行和列，为避免越界
			for (int j = 1; j < height; j++) {
				int x = j * width + i;
				int r = (pix[x] >> 16) & 0xff;
				int g = (pix[x] >> 8) & 0xff;
				int b = pix[x] & 0xff;
				pixelGray = (int) (0.3 * r + 0.59 * g + 0.11 * b);// 计算每个坐标点的灰度
				gray[i][j] = (pixelGray << 16) + (pixelGray << 8) + (pixelGray);
				graysum += pixelGray;
			}
		}
		graymean = (int) (graysum / area);// 整个图的灰度平均值
		average = graymean;
Log.i(TAG,"Average:"+average);
		for (int i = 0; i < width; i++) // 计算整个图的二值化阈值
		{
			for (int j = 0; j < height; j++) {
				if (((gray[i][j]) & (0x0000ff)) < graymean) {
					graybackmean += ((gray[i][j]) & (0x0000ff));
					back++;
				} else {
					grayfrontmean += ((gray[i][j]) & (0x0000ff));
					front++;
				}
			}
		}
		int frontvalue = (int) (grayfrontmean / front);// 前景中心
		int backvalue = (int) (graybackmean / back);// 背景中心
		float G[] = new float[frontvalue - backvalue + 1];// 方差数组
		int s = 0;
Log.i(TAG,"Front:"+front+"**Frontvalue:"+frontvalue+"**Backvalue:"+backvalue);
		for (int i1 = backvalue; i1 < frontvalue + 1; i1++)// 以前景中心和背景中心为区间采用大津法算法（OTSU算法）
		{
			back = 0;
			front = 0;
			grayfrontmean = 0;
			graybackmean = 0;
			for (int i = 0; i < width; i++) {
				for (int j = 0; j < height; j++) {
					if (((gray[i][j]) & (0x0000ff)) < (i1 + 1)) {
						graybackmean += ((gray[i][j]) & (0x0000ff));
						back++;
					} else {
						grayfrontmean += ((gray[i][j]) & (0x0000ff));
						front++;
					}
				}
			}
			grayfrontmean = (int) (grayfrontmean / front);
			graybackmean = (int) (graybackmean / back);
			G[s] = (((float) back / area) * (graybackmean - average)
					* (graybackmean - average) + ((float) front / area)
					* (grayfrontmean - average) * (grayfrontmean - average));
			s++;
		}
		float max = G[0];
		int index = 0;
		for (int i = 1; i < frontvalue - backvalue + 1; i++) {
			if (max < G[i]) {
				max = G[i];
				index = i;
			}
		}


		for (int i = 0; i < width; i++) {
			for (int j = 0; j < height; j++) {
				int in = j * width + i;
				if (((gray[i][j]) & (0x0000ff)) < (index + backvalue)) {
					pix[in] = Color.rgb(0, 0, 0);
				} else {
					pix[in] = Color.rgb(255, 255, 255);
				}
			}
		}
		
		Bitmap temp = Bitmap.createBitmap(width, height, Bitmap.Config.ARGB_8888);
		temp.setPixels(pix, 0, width, 0, 0, width, height);
		image.setImageBitmap(temp);
    }