Poor lighting differentiation between chip & background

[riversn]

Low lighting sometimes seems to work best on some samples to differentiate between the chip being present and the background.

For instance, the chip and the background will both 'turn blue' (register as color entropy) at the same threshold values when a user can clearly see a difference still.

It's hard to describe the issue well, but the SEYR color analysis is sometimes able to differentiate between when it's plotting Blue vs Purple in the threshold map, but it'd be better if maybe the regions are separated a bit more somehow.

Very hard to send a sample image set for you to see what we mean.

[riversn]

I liked the magic wand button, kinda sad it went away

[riversn]

Also, showed this to some of the coworkers and they were like, "Wow" "Oh my gosh" "We need this ASAP" "It's so cool"

[riversn]

If user selects Red chroma (same as default, but actually selecting it) it should fill red to show, not same as background like default does

[bradmartin333]

I'm thinking I will create a new wizard window with a teaser v1.4 format that will replace the chroma button and threshold slider as well as show the steps it takes to get the score. Might take a bit longer, so let me know if you'd rather I just make documentation about what it does. I can do the manual red thing, but it will mean an additional serialized boolean for each feature.

[riversn]

Whatever doesn't cause issues with performance probably.

I actually like the chroma button a lot - as it stand it really does work really well. Maybe the magic wand button could be replaced with a wizard hat and that could open the 'color wizard'. But leaving the chroma button as is.

[bradmartin333]

The chroma color both emphasizes channels and masks them. I use it to remove the cloudy background seen here:

That seems to give cleaner results.

I will hold off on the wizard then and just add the red color change then make a new demo .exe

[riversn]

I've got great pass/fail results with v1.3.28

White/Grey/DarkGrey seems to work the best across all samples. And for other samples, gold/orange, and light green, purple/blue, and red all have good/interesting results.

I do need a definition of what the color and scoring means. In the threshold map:

• black = empty
• white (and any color except purple or green?) = filled.

But fundamentally, this works amazingly well! I really can't be more excited with it in how much it helps improve yield fidelity.

[bradmartin333]

ok - I will create a graphic showing how it works

[bradmartin333]

I found some bugs in the scoring method - so expect slightly different, but more understandable results from changing the chroma

[bradmartin333]

infographic added to readme