opencv入门笔记之四 读写视频 捕捉移动点

opencv入门笔记之四 读写视频 捕捉移动点

 

 捕捉移动点时,  我们可以使用一种叫做 光学流动(optical flow) 的算法,  在关键点的周围,  所有点的流动导数都是相同的,  从而我门可以判断出关键点( feature points)

 下面是代码:

classCameraCaliberator

{

vector<vector<Point3f>> objectPoints;

// the real points

vector<vector<Point2f>> imagePoints;

// points found inimage

Mat cameraMatrix;

Mat distCoeffs;

int flag;

Mat map1, map2;

bool mustInitUndistort;

 

public:

CameraCaliberator() :flag(0), mustInitUndistort(true)

{

};

 

int addChessboardPoints(

const vector<string>& filelist, Size &boardSize){

vector<Point2f> imageCorners;

vector<Point3f> objectCorners;

for (int i = 0; i < boardSize.height;i++)

{

for (int j = 0; j < boardSize.width;j++)

{

objectCorners.push_back(Point3f(i, j,0.0f));        

}

}

Mat image;

int successes = 0;

for (int i = 0; i < filelist.size();i++)

{

image = imread(filelist[i], 0);

 

bool found = findChessboardCorners(image, boardSize, imageCorners);

cornerSubPix(image, imageCorners, Size(5, 5), Size(-1, -1),

TermCriteria(TermCriteria::MAX_ITER + TermCriteria::EPS, 30, 0.1));

//used to define when to stop

if (imageCorners.size() == boardSize.area())

{

addPoints(imageCorners, objectCorners);

successes++;

}

}

return successes;

}

double calibrate(Size &imageSize){

mustInitUndistort = true;

vector<Mat> rvecs, tvecs;

return        calibrateCamera(objectPoints,

imagePoints,

imageSize,

cameraMatrix,

distCoeffs,

rvecs,

tvecs,

flag);

}

Mat remap(const        Mat&image){

Mat undistorted;

if (mustInitUndistort){

initUndistortRectifyMap(cameraMatrix,

distCoeffs,

Mat(), Mat(),

image.size(),

CV_32FC1,

map1, map2);

 

}

mustInitUndistort = false;

cv::remap(image, undistorted, map1, map2, INTER_LINEAR);

return undistorted;

};

//        // Set the calibration options

//        // 8radialCoeffEnabled shouldbe true if 8 radial coefficients are required (5 is default)

//        // tangentialParamEnabledshould be true if tangeantial distortion is present

//        voidCameraCaliberator::setCalibrationFlag(bool radial8CoeffEnabled, booltangentialParamEnabled) {

//

//                //Set the flag used in cv::calibrateCamera()

//                flag= 0;

//                if(!tangentialParamEnabled) flag += CV_CALIB_ZERO_TANGENT_DIST;

//                if(radial8CoeffEnabled) flag += CV_CALIB_RATIONAL_MODEL;

//        }

 

 

private:

void addPoints(vector<Point2f>& imageCorners,vector<Point3f>& objectCorners){

imagePoints.push_back(imageCorners);

objectPoints.push_back(objectCorners);

}

};

 

 

 

void m(){

vector<string> filelist;

Size boardSize(6, 4);

for (int i = 1; i <= 20; i++) {

 

stringstream str;

str << "chessboards\\chessboard" <<std::setw(2) << std::setfill('0') << i << ".jpg";

cout << str.str() << endl;

 

filelist.push_back(str.str());

//                image= cv::imread(str.str(), 0);

//                cv::imshow("Image",image);

//

//                cv::waitKey(100);

}

Mat image=imread(filelist[0]);

CameraCaliberator cameraCaliberator;

 

cameraCaliberator.addChessboardPoints(filelist,boardSize);

cameraCaliberator.calibrate(image.size());

Matuimage=        cameraCaliberator.remap(image);

imshow(" ", uimage);

}

 

 

 

// Openthe video file

cv::VideoCapturecapture("bike.avi");

// checkif video successfully opened

if(!capture.isOpened())

return 1;

// Getthe frame rate

doublerate = capture.get(CV_CAP_PROP_FPS);

boolstop(false);

cv::Matframe; // current video frame

cv::namedWindow("ExtractedFrame");

// Delaybetween each frame in ms

//corresponds to video frame rate

int delay= 1000 / rate;

// forall frames in video

while(!stop) {

// read next frame if any

if (!capture.read(frame))

break;

cv::imshow("Extracted Frame", frame);

// introduce a delay

// or press key to stop

if (cv::waitKey(delay) >= 0)

stop = true;

}

// Closethe video file.

// Notrequired since called by destructor

capture.release();

 

 

 

class FeatureTracker : public FrameProcessor {

 

cv::Matgray;                        //current gray-level image

cv::Matgray_prev;                //previous gray-level image

std::vector<cv::Point2f> points[2]; // tracked features from0->1

std::vector<cv::Point2f> initial;   // initial position of tracked points

std::vector<cv::Point2f> features;  // detected features

int max_count;          // maximum number of features to detect

double qlevel;    // qualitylevel for feature detection

double minDist;   // minimumdistance between two feature points

std::vector<uchar> status; // status of tracked features

    std::vector<float> err;    // error in tracking

 

  public:

 

FeatureTracker() : max_count(500), qlevel(0.01), minDist(10.) {}

 

// processing method

void process(cv:: Mat &frame, cv:: Mat &output) {

 

// convert to gray-level image

cv::cvtColor(frame, gray, CV_BGR2GRAY);

frame.copyTo(output);

 

// 1. if new feature points must be added

if(addNewPoints())

{

// detect feature points

detectFeaturePoints();

// add the detected features to the currently tracked features

points[0].insert(points[0].end(),features.begin(),features.end());

initial.insert(initial.end(),features.begin(),features.end());

}

 

// for first image of the sequence

if(gray_prev.empty())

           gray.copyTo(gray_prev);

           

// 2. track features

cv::calcOpticalFlowPyrLK(gray_prev, gray, // 2 consecutive images

points[0], // input point position in first image

points[1], // output point postion in the second image

status,    // tracking success

err);      // tracking error

          

// 2. loop over the tracked points to reject the undesirables

int k=0;

for( int i= 0; i < points[1].size(); i++ ) {

 

// do we keep this point?

if (acceptTrackedPoint(i)) {

 

// keep this point in vector

initial[k]= initial[i];

points[1][k++] = points[1][i];

}

}

 

// eliminate unsuccesful points

        points[1].resize(k);

initial.resize(k);

 

// 3. handle the accepted tracked points

handleTrackedPoints(frame, output);

 

// 4. current points and image become previous ones

std::swap(points[1], points[0]);

        cv::swap(gray_prev, gray);

}

 

// feature point detection

void detectFeaturePoints() {

 

// detect the features

cv::goodFeaturesToTrack(gray, // the image

features,   // the outputdetected features

max_count,  // the maximumnumber of features

qlevel,     // quality level

minDist);   // min distancebetween two features

}

 

// determine if new points should be added

bool addNewPoints() {

 

// if too few points

return points[0].size()<=10;

}

 

// determine which tracked point should be accepted

bool acceptTrackedPoint(int i) {

 

return status[i] &&

// if point has moved

(abs(points[0][i].x-points[1][i].x)+

(abs(points[0][i].y-points[1][i].y))>2);

}

 

// handle the currently tracked points

void handleTrackedPoints(cv:: Mat &frame, cv:: Mat &output){

 

// for all tracked points

for(int i= 0; i < points[1].size(); i++ ) {

 

// draw line and circle

    cv::line(output,initial[i], points[1][i], cv::Scalar(255,255,255));

cv::circle(output, points[1][i], 3, cv::Scalar(255,255,255),-1);

}

}

};

 

#endif

 

posted on 2015-02-08 21:55 rex686568 阅读( ...) 评论( ...) 编辑 收藏

转载于:https://www.cnblogs.com/Pomodori/p/4316618.html