1.1. img_prep()

This function is used if you want to find contours in the image within a certain color range.

Usage

img_prep(img, low, high, iterations, blur_strength)
Parameters:

  • img: image to be processed. Must be a numpy array;

  • low: lower color range in HSV color space. Default is tuple (0, 0, 0). Must be a tuple of 3 integers;

  • high: upper color range in HSV color space. Default is tuple (255, 255, 255). Must be a tuple of 3 integers;

  • iterations: number of iterations for the morphological transformation. Default is 1. Must be an integer;

  • blur_strength: strength of the blur. Default is 0. Must be an integer.

Returns:

  • dilated: processed image

  • contours: list of contours found in the image

  • hierarchy: list of hierarchy found in the image

Examples

If you input only picture, then it will completely satisfy the HSV range [0, 0, 0] - [255, 255, 255]:

transform.img_prep(frame)
Full HSV range

By changing the range of the frame for the filter, you can choose which part of the image remains:

transform.img_prep(frame, (100, 0, 0), (255, 255, 255))
transform.img_prep(frame, (0, 100, 100), (155, 255, 255))
Part of the HSV range

By increasing the blur strength, you can achieve smoother borders and remove noise from the image:

transform.img_prep(frame, (1, 0, 0), (255, 255, 255), 0, 0)
transform.img_prep(frame, (1, 0, 0), (255, 255, 255), 0, 10)
Blur strength

By increasing the number of iterations, you can close badly visible contours

transform.img_prep(frame, (1, 0, 0), (255, 255, 255), 0, 0)
transform.img_prep(frame, (1, 0, 0), (255, 255, 255), 1, 0)
transform.img_prep(frame, (1, 0, 0), (255, 255, 255), 5, 0)
Iterations