Hidden Colors - with a new feature

[DavidWright2010]

My abstracts are almost always created from a single image, using my (top-secret) processing method. (I have identified some few images as composites.) However, my process sometimes result in non-negligible black patches, which IMO suck the energy out of the resultant image.

So I have created a process (which I call 'mitochrome'), wherein I replace some black patches - which meet a certain criterium - with the corresponding regions (and colors) from another processed image.

LHS is the original 'hidden colors' image; RHS has the same image with the mitochrome process applied. (Notice that the LHS image has a few naturally occurring mitochromes.) What do you think - better or not?

David

 

[Brian]

It looks like the original Black spot underflowed (or went negative?) a look-up-table of some type. The new image- (just guessing) looks like you use the same (or close) LUT for the values in the underflowed area. It does make it look like a Fractal.

I'll have to dig out my M8 DNG processing code, uses the color values of each Bayer Cell as real/imaginary components to compute the Mandelbrot set. The image retains spatial features, with some hard-to-predict color values.

The image reminds me of the microscope slides of cells being attacked by cancer my wife generated a long time ago. We developed some image processing techniques to identify the cells, lots of spatial features extracted. The color scheme used reminds me of this.

 

[DavidWright2010]

It looks like the original Black spot underflowed (or went negative?) a look-up-table of some type. The new image- (just guessing) looks like you use the same (or close) LUT for the values in the underflowed area. It does make it look like a Fractal.

That's a possibility.

The 'mitochromes' look similar to the original image because they were obtained by the same method. That image lacked any large-scale structure and was less interesting to me. But a good source of mitochromes.

David