Revision history [back]

Hi there!

I am using the calibrateHandEye functions from the Calib3d library to get the handEye calibration of a UR10 robot with a camera RealSense fixed to the end effector. After some problems I think, that I got to get the function work properly. But now I am seeing two things that do not fit completely, and I was hoping for a third opinion from your point of view.

As you can see from the image below, I made 4 four measurements for each of the 5 methods that the function proposes. For each of the methods I used 10, 15, 18 and 20 robot poses with their respective pictures. From here I got a little confused:

1. The first method (CALIB_HAND_EYE_TSAI) give completely different values depending on the amount of robot poses used. At the same time, and with the same values, the others 4 methods seem to converge to certain values close enough between them. Why is that?
2. Though the values of the last 4 methods seem to converge, between them there are important differences. Why is that?

Thank you for the help!

Hi there!

I am using the calibrateHandEye functions from the Calib3d library to get the handEye calibration of a UR10 robot with a camera RealSense fixed to the end effector. After some problems I think, that I got to get the function work properly. But now I am seeing two things that do not fit completely, and I was hoping for a third opinion from your point of view.

As you can see from the image below, I made 4 four measurements for each of the 5 methods that the function proposes. For each of the methods I used 10, 15, 18 and 20 robot poses with their respective pictures. From here I got a little confused:

1. The first method (CALIB_HAND_EYE_TSAI) give completely different values depending on the amount of robot poses used. At the same time, and with the same values, the others 4 methods seem to converge to certain values close enough between them. Why is that?
2. Though the values of the last 4 methods seem to converge, between them there are important differences. Why is that?

Thank you for the help!

Hi there!

I am using the calibrateHandEye functions from the Calib3d library to get the handEye calibration of a UR10 robot with a camera RealSense fixed to the end effector. After some problems I think, that I got to get the function work properly. But now I am seeing two things that do not fit completely, and I was hoping for a third opinion from your point of view.

As you can see from the image below, I made 4 four measurements for each of the 5 methods that the function proposes. For each of the methods I used 10, 15, 18 and 20 robot poses with their respective pictures. From here I got a little confused:

1. The first method (CALIB_HAND_EYE_TSAI) give completely different values depending on the amount of robot poses used. At the same time, and with the same values, the others 4 methods seem to converge to certain values close enough between them. Why is that?
2. Though the values of the last 4 methods seem to converge, between them there are important differences. Why is that?

Thank you for the help!

Link to get the images and poses.

Hi there!

I am using the calibrateHandEye functions from the Calib3d library to get the handEye calibration of a UR10 robot with a camera RealSense fixed to the end effector. After some problems I think, that I got to get the function work properly. But now I am seeing two things that do not fit completely, and I was hoping for a third opinion from your point of view.

As you can see from the image below, I made 4 four measurements for each of the 5 methods that the function proposes. For each of the methods I used 10, 15, 18 and 20 robot poses with their respective pictures. From here I got a little confused:

1. The first method (CALIB_HAND_EYE_TSAI) give completely different values depending on the amount of robot poses used. At the same time, and with the same values, the others 4 methods seem to converge to certain values close enough between them. Why is that?
2. Though the values of the last 4 methods seem to converge, between them there are important differences. Why is that?

Thank you for the help!

The camera intrinsic parameters are:

std::vector<double> distortionCoefficients(5);  // camera distortion
distortionCoefficients[0] = 9.6349551984637724e-02;
distortionCoefficients[1] = -3.3260675111130217e-01;
distortionCoefficients[2] = 0;
distortionCoefficients[3] = 0;
distortionCoefficients[4] = 2.5833277679122602e-01;

double f_x = 1.2993539019658076e+03; // Focal length in x axis
double f_y = 1.2993539019658076e+03; // Focal length in y axis (usually the same?)
double c_x = 960; // Camera primary point x
double c_y = 540; // Camera primary point y

cv::Mat cameraMatrix(3, 3, CV_32FC1);
cameraMatrix.at<float>(0, 0) = f_x;
cameraMatrix.at<float>(0, 1) = 0.0;
cameraMatrix.at<float>(0, 2) = c_x;
cameraMatrix.at<float>(1, 0) = 0.0;
cameraMatrix.at<float>(1, 1) = f_y;
cameraMatrix.at<float>(1, 2) = c_y;
cameraMatrix.at<float>(2, 0) = 0.0;
cameraMatrix.at<float>(2, 1) = 0.0;
cameraMatrix.at<float>(2, 2) = 1.0;

Link to get the images and poses.