Color Transforms Module - Image Engineering camSPECS User Manual

3. Select the ColorChecker reflectance data. This data contains the spectral distribution between 380 and
730 nm in 5 nm increments.
4. Select an RGB image file from a ColorChecker test chart (ColorChecker with 24 patches or Color
Checker SG with 140 patches). The image must be a raw image file. Do not use compressed image
formats like JPEG or other output-referred encodings since source colors have been rendered.
5. Select a dark image for subtracting from the ColorChecker image (optional). The size of this image and
the ColorChecker image must be the same.
6. After clicking the Start button, the selected ColorChecker image is opened and displayed. Select the
chart by drawing a rectangle around the patches from upper left to lower right and double-click on the
rectangle. The ROIs for each patch are displayed in the next step and can be rearranged manually.
Double-click outside ROIs to continue.
7. Export the validation graph to an image (JPEG, PNG, TIFF, BMP, GIF) for documentation purposes.
The validation verifies the processing chain from the exposure of the ColorChecker images (surrounding light
conditions, camera lenses, raw image converting, etc.) to the spectral response measurement with
camSPECS. In the displayed result, each patch's measured digital values (red, green, and blue) are plotted
against the predicted digital values. All values are normalized to a range between 0 and 1. If the exposure
and measurement setup is highly accurate, all RGB data points in this plot should be close to the white
diagonal line (best correlation). Possible reasons for significant variations may be changes in the illumination
of the ColorChecker or manipulations of the RGB data within the camera.
4.8

Color transforms module

The color transforms module computes various color transforms from the measured camera spectral
sensitivities for several different types of transforms.
Image Engineering
Page 26 of 36