我是一个有用的助手,可以为你进行翻译。以下是需要翻译的内容:
我尝试在Java的OpenCV程序中实现卡尔曼滤波器。 我对OpenCV和Kalman滤波器都很新。 我在C ++中找到了一些示例(Java中没有太多),这是我迄今为止所拥有的:
初始化:
//create kalman filter
KalmanFilter kalman = new KalmanFilter(4,2,0,CvType.CV_32F);
//set transition matrix
float[] tM = { 1, 0, 1, 0,
0, 1, 0, 1,
0, 0, 1, 0,
0, 0, 0, 1 } ;
Mat transitionMatrix=new Mat(4,4,CvType.CV_32F,new Scalar(0));
transitionMatrix.put(0,0,tM);
kalman.set_transitionMatrix(transitionMatrix);
//set init measurement
Mat measurementMatrix = new Mat (2,1, CvType.CV_32F);
measurementMatrix.setTo(new Scalar(0));
kalman.set_measurementMatrix(measurementMatrix);
//Set state matrix
Mat statePre = new Mat(4,1, CvType.CV_32F);
statePre.put(1, 1, 300);
statePre.put(2, 1, 200);
statePre.put(3, 1, 0);
statePre.put(4, 1, 0);
kalman.set_statePre(statePre);
//Process noise Covariance matrix
Mat processNoiseCov=Mat.eye(4,4,CvType.CV_32F);
processNoiseCov=processNoiseCov.mul(processNoiseCov,1e-1);
kalman.set_processNoiseCov(processNoiseCov);
//Measurement noise Covariance matrix: reliability on our first measurement
Mat measurementNoiseCov=Mat.eye(4,4,CvType.CV_32F);
measurementNoiseCov=measurementNoiseCov.mul(measurementNoiseCov,1e-1);
kalman.set_measurementNoiseCov(measurementNoiseCov);
Mat id2=Mat.eye(4,4,CvType.CV_32F);
id2=id2.mul(id2,0.1);
kalman.set_errorCovPost(id2);
针对每一帧视频:
prediction= kalman.predict();
predictPt.x = prediction.get(1,1)[0];
predictPt.y = prediction.get(2,1)[0];
...新的测量方式...
measurementMatrix.put(1, 1, center.x);
measurementMatrix.put(2, 1, center.y);
measPt.x=center.x;
measPt.y=center.y;
Mat estimated = kalman.correct(measurementMatrix);
statePt.x=estimated.get(1, 1)[1];
statePt.y= estimated.get(2, 1)[1];
问题是我得到了一个空的预测值,但我不知道为什么会出现这种情况。有人知道我的代码有哪些问题吗?非常感谢任何帮助!
谢谢!