Processing of colour images

The images of NGC891 and M65/66 are assembled from MYC- and black/white images in the following way: On day 1 exposures are made without colour filters. After adding the images and processing with an unsharp masking technique the image L results. On day 2 exposures are made using three colour filters M (Magenta), Y (Yellow) and C (Cyan), combined with an IR-blocking filter. The exposures are added and the resulting M-, Y- and C-images are converted to RGB images using the next formula (pixel by pixel):

R=Y+M-C
G=Y+C-M
B=C+M-Y

A possible background in these images is subtracted and the images are balanced by multiplying the images by certain constants. These constants are determined using exposures of the moon for which the R, G and B-components should be nearly equal. For the MYC filter set I use, and the FTT1010-M CCD chip the following multiplication constants are used:

R: 1.07
G: 1.00 (by definition)
B: 2.26

Next, the images are converted to three images Y, I and Q using the following matrix-conversion:

Y=0.299*R+0.587*G+0.114*B
I=0.596*R-0.275*G-0.321*B
Q=0.212*R-0.523*G+0.311*B

The Y image now contains the intensity component of the colour image, while the images I and Q contain the colour information. In general, due to the use of colour filters resulting in lower signals, the quality of the Y-image is less than of the black/white image obtained on day 1. The Y-image therefore is exchanged for the L-image, and the YIQ-images are converted to the RGB-images using the inverse matrix conversion:

R=Y+0.956*I+0.621*Q
G=Y-0.272*I-0.647*Q
B=Y-1.105*I+1.702*Q

Finally, these images are combined to a colour image in .BMP format and compressed to a Jpeg-format.