问题描述

opencv中的kmeans函数

LBPimage是一个3通道的灰度图像。通过kmeans函数之后，分的两类，labels的各个分量都是0，只有第一个和最后一个元素是1.请问大神们，错在哪里。

Mat LBPimage=LBP(imgs);
int nl = LBPimage.rows;
int nc = LBPimage.cols;
Mat points;
points.create(nc*nl, 1, CV_32FC1);
int n = 0;
for (int i = 0; i < nl; i++)
{
    for (int j = 0; j < nc; j++)
    {
        points.at<int>(n,0) = (int)LBPimage.at<Vec3b>(i, j)[0];
        n++;
    }
}
Mat labels;
Mat centers;
kmeans(points, 2, labels, TermCriteria(CV_TERMCRIT_EPS,10000,0.01), 50, KMEANS_PP_CENTERS, centers);