CHROM Python API

bob.rppg.chrom.extract_utils.compute_mean_rgb(image, mask=None)

computes the mean R, G and B of an image.

Note that a mask could be provided to tell which pixels should be taken into account when computing the mean.

Parameters:

• image (numpy.ndarray) – The image to process
• mask (numpy.ndarray) – Mask of the size of the image, telling which pixels should be considered

Returns:

• mean_r (float) – The mean red value
• mean_g (float) – The mean green value
• mean_b (float) – The mean blue value

bob.rppg.chrom.extract_utils.compute_gray_diff(previous, current)

computes the difference in intensity between two images.

Parameters:

• previous (numpy.ndarray) – The previous frame.
• current (numpy.ndarray) – The current frame.

Returns:

The sum of the absolute difference in pixel intensity between two frames

Return type:

float

bob.rppg.chrom.extract_utils.select_stable_frames(diff, n)

selects a stable subset of consecutive frames

The selection is made by considering the grayscale difference between frames. The subset is chosen as the one for which the sum of difference is minimized

Parameters:

• diff (numpy.ndarray) – The sum of absolute pixel intensity differences between consecutive frames, across the whole sequence.
• n (int) – The number of consecutive frames you want to select.

Returns:

index – The frame index at which the stable segment begins.

Return type:

int

bob.rppg.chrom.extract_utils.project_chrominance(r, g, b)

Projects rgb values onto the x and y chrominance space

See equation (9) of [dehaan-tbe-2013].

Parameters:

• r (float) – The red value
• g (float) – The green value
• b (float) – The blue value

Returns:

• x (float) – The x value
• y (float) – The y value