在许多实际运用中,我们需要分割图像,但无法从背景图像中获得有用信息。分水岭算法(watershed algorithm)在这方面往往是非常有效的。此算法可以将图像中的边缘转化为“山脉”,将均匀区域转化为“山谷”,这样有助于分割目标。
该算法是一种基于拓扑理论的数学形态学的分割方法,其基本思想是把图像看作测地学上的拓扑地貌,图像中每一点像素的灰度值表示该点的海拔高度,每一个局部极小值及其影响区域称为集水盆,而集水盆的边界则形成分水岭。
在把图像传给函数之前,我们需要大致勾画标记出图像中的期望进行分割的区域,它们被标记为正指数。所以,每一个区域都会被标记为像素值1、2、3等,表示成为一个或者多个连接组件。这些标记值可以使用findContours函数和drawContours函数由二进制的掩码检索出来。
void watershed(InputArray image, InputOutputArray markers)
#include
#include
#include
#include
using namespace cv;
using namespace std;
#define WINDOW_NAME1 "【程序窗口1】"
#define WINDOW_NAME2 "【分水岭算法效果图】"
Mat g_maskImage, g_srcImage;
Point prevPt(-1, -1);
static void ShowHelpText();
static void on_Mouse(int event, int x, int y, int flags, void*);
int main(int argc, char** argv)
{
system("color 6F");
ShowHelpText();
g_srcImage = imread("fg.jpg", 1);
imshow(WINDOW_NAME1, g_srcImage);
Mat srcImage, grayImage;
g_srcImage.copyTo(srcImage);
cvtColor(g_srcImage, g_maskImage, COLOR_BGR2GRAY);
cvtColor(g_maskImage, grayImage, COLOR_GRAY2BGR);
g_maskImage = Scalar::all(0);
setMouseCallback(WINDOW_NAME1, on_Mouse, 0);
while (1)
{
int c = waitKey(0);
if ((char)c == 27)
break;
if ((char)c == '2')
{
g_maskImage = Scalar::all(0);
srcImage.copyTo(g_srcImage);
imshow("image", g_srcImage);
}
if ((char)c == '1' || (char)c == ' ')
{
int i, j, compCount = 0;
vector<vector<Point> > contours;
vector<Vec4i> hierarchy;
findContours(g_maskImage, contours, hierarchy, CV_RETR_CCOMP, CV_CHAIN_APPROX_SIMPLE);
if (contours.empty())
continue;
Mat maskImage(g_maskImage.size(), CV_32S);
maskImage = Scalar::all(0);
for (int index = 0; index >= 0; index = hierarchy[index][0], compCount++)
drawContours(maskImage, contours, index, Scalar::all(compCount + 1), -1, 8, hierarchy, INT_MAX);
if (compCount == 0)
continue;
vector<Vec3b> colorTab;
for (i = 0; i < compCount; i++)
{
int b = theRNG().uniform(0, 255);
int g = theRNG().uniform(0, 255);
int r = theRNG().uniform(0, 255);
colorTab.push_back(Vec3b((uchar)b, (uchar)g, (uchar)r));
}
double dTime = (double)getTickCount();
watershed(srcImage, maskImage);
dTime = (double)getTickCount() - dTime;
printf("\t处理时间 = %gms\n", dTime * 1000. / getTickFrequency());
Mat watershedImage(maskImage.size(), CV_8UC3);
for (i = 0; i < maskImage.rows; i++)
for (j = 0; j < maskImage.cols; j++)
{
int index = maskImage.at<int>(i, j);
if (index == -1)
watershedImage.at<Vec3b>(i, j) = Vec3b(255, 255, 255);
else if (index <= 0 || index > compCount)
watershedImage.at<Vec3b>(i, j) = Vec3b(0, 0, 0);
else
watershedImage.at<Vec3b>(i, j) = colorTab[index - 1];
}
watershedImage = watershedImage * 0.5 + grayImage * 0.5;
imshow(WINDOW_NAME2, watershedImage);
}
}
return 0;
}
static void on_Mouse(int event, int x, int y, int flags, void*)
{
if (x < 0 || x >= g_srcImage.cols || y < 0 || y >= g_srcImage.rows)
return;
if (event == CV_EVENT_LBUTTONUP || !(flags & CV_EVENT_FLAG_LBUTTON))
prevPt = Point(-1, -1);
else if (event == CV_EVENT_LBUTTONDOWN)
prevPt = Point(x, y);
else if (event == CV_EVENT_MOUSEMOVE && (flags & CV_EVENT_FLAG_LBUTTON))
{
Point pt(x, y);
if (prevPt.x < 0)
prevPt = pt;
line(g_maskImage, prevPt, pt, Scalar::all(255), 5, 8, 0);
line(g_srcImage, prevPt, pt, Scalar::all(255), 5, 8, 0);
prevPt = pt;
imshow(WINDOW_NAME1, g_srcImage);
}
}
static void ShowHelpText()
{
printf("\n\n ----------------------------------------------------------------------------\n");
printf("\t请先用鼠标在图片窗口中标记出大致的区域,\n\n\t然后再按键【1】或者【SPACE】启动算法。"
"\n\n\t按键操作说明: \n\n"
"\t\t键盘按键【1】或者【SPACE】- 运行的分水岭分割算法\n"
"\t\t键盘按键【2】- 恢复原始图片\n"
"\t\t键盘按键【ESC】- 退出程序\n\n\n");
}
运行效果如下: