Camera - centric Point of image_geometry::PinholeCameraModel [closed]

edit

Hey guys,

I am trying to understand the project3dToPixel(...) function of image_geometry::PinholeCameraModel, but still not sure if I understand it right, because I get other results then I think I should get.....

Here is a piece of the code with which I try to get the camera centric Point:

cv::Point2d pInImage;   
cv::Point3d coordinates;
image_geometry::PinholeCameraModel cameraModel;

// Axis directions in the image plane are: x - to the right | y - down | z - into the plane
coordinates.x = 0;
coordinates.y = 0;
coordinates.z = 1;   

RCLCPP_INFO(this->get_logger(), "Coordinates: x=%f | y=%f | z=%f", coordinates.x, coordinates.y, coordinates.z);

RCLCPP_INFO(this->get_logger(), "-------------------------------------");

cameraModel.fromCameraInfo(cameraInfo);
pInImage = cameraModel.project3dToPixel(coordinates);

RCLCPP_INFO(this->get_logger(), "Rubik Cube in image: u=%f | v=%f", pInImage.x, pInImage.y);

RCLCPP_INFO(this->get_logger(), "-------------------------------------");

This is the calibration file of the camera which I use:

image_width: 640
image_height: 480
camera_name: trust_qhd_webcam:_trust_qhd_web
camera_matrix:
rows: 3
cols: 3
data: [626.00986,   0.     , 313.4817 ,
                   0.     , 626.95054, 254.95665,
                   0.     ,   0.     ,   1.     ]
distortion_model: plumb_bob
distortion_coefficients:
rows: 1
cols: 5
data: [0.146588, -0.316449, 0.005444, -0.001867, 0.000000]
rectification_matrix:
rows: 3
cols: 3
data: [1., 0., 0.,
          0., 1., 0.,
          0., 0., 1.] 
projection_matrix:
rows: 3
cols: 4
data: [636.19208,   0.     , 311.96673,   0.     ,
                   0.     , 636.69824, 256.23645,   0.     ,
                   0.     ,   0.     ,   1.     ,   0.     ]

Because the image plane of my camera has the size of 640x480 I would assume, that if I want to project a 3D Point with the coordinates x=0, y=0, z=1 I would get u=320, v=240 as the image plane coordinates, but I am getting u=311.966730, v=256.236450....So its almost right, but where does the difference from the expected result comes from?

Could this be an issue because of bad calibration of the camera?