I have made a test chart recognition template for the basICColor dcam target +.
The special feature of this target is a light trap -- a round hole punched in the middle of the target and a black box behind it. The hole is somewhat bigger than the other patches so that the adjacent fields are affected. I tried to account for this by making the adjacent fields smaller, but apparently it doesn't work. I will attach a downscaled version of the target, a detail of the diagnostic file (scanin -v -dionap) and my .cht file.
Did I make a mistake? Is it actually essential to omit the border of the hole or does scanin take out such irregularities (comparable to scratches or dust) automatically by some kind of weighted mean algorithm?
TIA, Klaus
