I've done lacunarity and fractal analysis before, as well as simple statistical analysis involving the RGB channels of color, but I am expecting that there are better ways. I've come across "color PCA," but the papers I've found are very highly mathematical (not my domain at all!). Does anyone know how to use that for this purpose?
If I understand the question correctly you want a software which will be able to identify species, right?
Yes, and for discriminating among species based on the design patterns.
And could be usefull compare between different means for a range of gray scale??? I mean...if you change the pictures to WB colors...and then get a histogram for gray scale values for each picture (with a mean for each pic)...in this case, image J is a powerfull tool in image processing...and has another options that maybe can help you.
If you are interested in the shape of the patterns you could use geometric morphometrics.
Are you sure that color instead of morphometric shell characters are the way to go for species discrimination in your case (I do not know which group you are interested in)? If you need help in morphometric analyses of shell shape, you probably want to contact Markus Pfenninger from the University of Frankfurt, he did considerable work on shell shape analysis for species discrimination.
Best
Carsten
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology