trifinger_cameras.utils module

Utility functions.

trifinger_cameras.utils.check_image_sharpness(image: ndarray, canny_threshold1: float = 25.0, canny_threshold2: float = 250.0) → Tuple[float, ndarray]

Estimate sharpness of the given image, using edge detection.

Uses Canny edge detection to estimate how sharp the images are (more edges = sharper). If the mean value of the edge image is below a certain threshold, this might mean that the corresponding camera is out of focus and should be checked.

See https://stackoverflow.com/a/66557408

Parameters:

• image – Input image.

• canny_threshold1 – See cv2.Canny.

• canny_threshold2 – See cv2.Canny.

Returns:

Tuple (edge_mean, edge_image). Where edge_mean is the mean value of the edge image. A higher mean value means more edges and thus indicates a sharper image. edge_image shows the detected edges. It is returned mostly for debugging and visualisation purposes.

trifinger_cameras.utils.convert_image(raw_image, format: str = 'bgr') → ndarray

Convert raw image from camera observation.

Parameters:

• raw_image – Raw image from camera observation.

• format (str) – Format of the output image. One of “bgr”, “rgb”, “gray”. Defaults to “bgr” which is the default format of OpenCV.

Returns:

The converted image as NumPy array.

trifinger_cameras.utils.print_tricamera_sensor_info(tricamera_info: TriCameraInfo) → None

Pretty-print the sensor info struct of the TriCamera driver.

trifinger_cameras.utils.rodrigues_to_matrix(rvec)

Convert Rodrigues vector to homogeneous transformation matrix.

Parameters:

rvec (array-like) – Rotation in Rodrigues format as returned by OpenCV.

Returns:

Given rotation as a 4x4 homogeneous transformation matrix.

Return type:

quaternion (array-like)