原文:
http://blog.csdn.net/zhaocj/article/details/45198965
DenseFeatureDetector可以生成具有一定密度和规律分布的图像特征点,它主要用于3D VIZ。
DenseFeatureDetector的原理是,把输入图像分割成大小相等的网格,每一个网格提取一个像素作为特征点。类似于图像尺度金字塔,该方法也可以生成不同层图像的特征点,每一层图像所分割的网格大小是不同的,即表示各层的尺度不同。
下面我们就来分析它的源码。
DenseFeatureDetector类的构造函数:
[cpp] view plain copy print?
- DenseFeatureDetector::DenseFeatureDetector( float _initFeatureScale, int _featureScaleLevels,
- float _featureScaleMul, int _initXyStep,
- int _initImgBound, bool _varyXyStepWithScale,
- bool _varyImgBoundWithScale ) :
- initFeatureScale(_initFeatureScale), featureScaleLevels(_featureScaleLevels),
- featureScaleMul(_featureScaleMul), initXyStep(_initXyStep), initImgBound(_initImgBound),
- varyXyStepWithScale(_varyXyStepWithScale), varyImgBoundWithScale(_varyImgBoundWithScale)
- {}
DenseFeatureDetector::DenseFeatureDetector( float _initFeatureScale, int _featureScaleLevels,
float _featureScaleMul, int _initXyStep,
int _initImgBound, bool _varyXyStepWithScale,
bool _varyImgBoundWithScale ) :
initFeatureScale(_initFeatureScale), featureScaleLevels(_featureScaleLevels),
featureScaleMul(_featureScaleMul), initXyStep(_initXyStep), initImgBound(_initImgBound),
varyXyStepWithScale(_varyXyStepWithScale), varyImgBoundWithScale(_varyImgBoundWithScale)
{}
initFeatureScale表示初始图像层特征点的尺度,默认为1
featureScaleLevels表示需要构建多少层图像,默认为1
featureScaleMul表示各层图像之间参数的比例系数,该系数等于相邻两层图像之间的网格宽度之比,尺度之比,以及预留边界宽度之比,默认为0.1
initXyStep表示初始图像层的网格宽度,默认为6
initImgBound表示初始图像层的预留边界宽度,默认为0
varyXyStepWithScale表示各层图像是否进行网格宽度的调整,如果为false,则表示各层图像网格宽度都是initXyStep,如果为true,则表示各层图像网格宽度不等,它们之间的比例系数为featureScaleMul,默认为true
varyImgBoundWithScale表示各层图像是否进行预留边界宽度的调整,如果为false,则表示各层图像预留边界宽度都是initImgBound,如果为true,则表示各层图像预留边界宽度不等,它们之间的比例系数为featureScaleMul,默认为false
下面是检测特征点函数detectImpl:
[cpp] view plain copy print?
- void DenseFeatureDetector::detectImpl( const Mat& image, vector<KeyPoint>& keypoints, const Mat& mask ) const
- {
- // curScale表示当前层图像特征点的尺度
- float curScale = static_cast<float>(initFeatureScale);
- //curStep表示当前层图像的网格宽度
- int curStep = initXyStep;
- //curBound表示当前层图像的预留边界宽度
- int curBound = initImgBound;
- //遍历各层图像
- for( int curLevel = 0; curLevel < featureScaleLevels; curLevel++ )
- {
- //遍历当前层图像的所有网格,图像四周的预留边界处是没有网格的,横、纵坐标的步长就是网格的宽度
- for( int x = curBound; x < image.cols - curBound; x += curStep )
- {
- for( int y = curBound; y < image.rows - curBound; y += curStep )
- {
- //把网格的左上角坐标处的像素作为该网格的特征点,并保存
- keypoints.push_back( KeyPoint(static_cast<float>(x), static_cast<float>(y), curScale) );
- }
- }
- //调整下一层图像特征点的尺度
- curScale = static_cast<float>(curScale * featureScaleMul);
- //如果varyXyStepWithScale为true,则调整下一层图像的网格宽度
- if( varyXyStepWithScale ) curStep = static_cast<int>( curStep * featureScaleMul + 0.5f );
- //如果varyImgBoundWithScale为true,则调整下一层图像的预留边界宽度
- if( varyImgBoundWithScale ) curBound = static_cast<int>( curBound * featureScaleMul + 0.5f );
- }
- //掩码矩阵的特征点处理
- KeyPointsFilter::runByPixelsMask( keypoints, mask );
- }
void DenseFeatureDetector::detectImpl( const Mat& image, vector<KeyPoint>& keypoints, const Mat& mask ) const
{
// curScale表示当前层图像特征点的尺度
float curScale = static_cast<float>(initFeatureScale);
//curStep表示当前层图像的网格宽度
int curStep = initXyStep;
//curBound表示当前层图像的预留边界宽度
int curBound = initImgBound;
//遍历各层图像
for( int curLevel = 0; curLevel < featureScaleLevels; curLevel++ )
{
//遍历当前层图像的所有网格,图像四周的预留边界处是没有网格的,横、纵坐标的步长就是网格的宽度
for( int x = curBound; x < image.cols - curBound; x += curStep )
{
for( int y = curBound; y < image.rows - curBound; y += curStep )
{
//把网格的左上角坐标处的像素作为该网格的特征点,并保存
keypoints.push_back( KeyPoint(static_cast<float>(x), static_cast<float>(y), curScale) );
}
}
//调整下一层图像特征点的尺度
curScale = static_cast<float>(curScale * featureScaleMul);
//如果varyXyStepWithScale为true,则调整下一层图像的网格宽度
if( varyXyStepWithScale ) curStep = static_cast<int>( curStep * featureScaleMul + 0.5f );
//如果varyImgBoundWithScale为true,则调整下一层图像的预留边界宽度
if( varyImgBoundWithScale ) curBound = static_cast<int>( curBound * featureScaleMul + 0.5f );
}
//掩码矩阵的特征点处理
KeyPointsFilter::runByPixelsMask( keypoints, mask );
}
下面给出一个具体的应用实例:
[cpp] view plain copy print?
- #include "opencv2/core/core.hpp"
- #include "highgui.h"
- #include "opencv2/imgproc/imgproc.hpp"
- #include "opencv2/features2d/features2d.hpp"
- #include "opencv2/nonfree/nonfree.hpp"
- using namespace cv;
- //using namespace std;
- int main(int argc, char** argv)
- {
- Mat img = imread("box_in_scene.png"), img1;;
- cvtColor( img, img1, CV_BGR2GRAY );
- DenseFeatureDetector dense;
- vector<KeyPoint> key_points;
- Mat output_img;
- dense.detect(img1,key_points,Mat());
- drawKeypoints(img, key_points, output_img, Scalar::all(-1), DrawMatchesFlags::DRAW_RICH_KEYPOINTS);
- namedWindow("DENSE");
- imshow("DENSE", output_img);
- waitKey(0);
- return 0;
- }