定义:FAST算法定义特征点是如果某个像素点和他周围领域足够多的像素点处于不同区域,那么这个像素点就是特征点。
对于灰度图,及特征点处灰度值与周围足够多像素灰度值不同。
tatic Ptr<FastFeatureDetector> create( int threshold=10, bool nonmaxSuppression=true,
int type=FastFeatureDetector::TYPE_9_16 );
threshold:边缘轨迹与中心点差值。
nonmaxSuppression:是否使用非极大值抑制(聚簇情况下使用)
FastFeatureDetector:
- TYPE_5_8 从轨迹中取8个点,当有5个点满足条件,就是特征点.
- TYPE_7_12 取轨迹12个点,7个满足条件,就是特征点.
- TYPE_9_16 取轨迹16个点,当9个满足条件,就是特征点
FASH:
无多尺度问题
计算速度快
健壮性差
//FASH
int main(int argc,char* argv[])
{
Mat srcImage = imread("FeatureDetectSrc1.jpg");
Mat srcGrayImage;
if (srcImage.channels() == 3)
{
cvtColor(srcImage,srcGrayImage,CV_RGB2GRAY);
}
else
{
srcImage.copyTo(srcGrayImage);
}
vector<KeyPoint>detectKeyPoint;
Mat keyPointImage1,keyPointImage2;
Ptr<FastFeatureDetector> fast = FastFeatureDetector::create();
fast->detect(srcGrayImage,detectKeyPoint);
drawKeypoints(srcImage,detectKeyPoint,keyPointImage1,Scalar(0,0,255),DrawMatchesFlags::DRAW_RICH_KEYPOINTS);
drawKeypoints(srcImage,detectKeyPoint,keyPointImage2,Scalar(0,0,255),DrawMatchesFlags::DEFAULT);
imshow("原",srcImage);
imshow("keyPoint——img1",keyPointImage1);
imshow("keyPoint——img2",keyPointImage2);
waitKey(0);
return 0;
}