Hello, I am trying to associate the color of an image to a single numerical value. Images (see attach) correspond to different aloe samples that change in color according to the degree of oxidation (green when there is no oxidation, brown for oxidated ones). I need to associate each image color to a single value since it is the outcome of a design of experiments to evaluate different conditions that can lead (or not) to different degrees of aloe oxidation. I am expecting different color tones for different degrees of oxidation.
I am new to this area, so all recommendation will be really helpful. Thanks!
Do I understand it right that you want to find some sort of calibration curve that will relate the colour to the degree of oxidation? I assume you have images in RGB colour space, which means each sample will be characterised by 3 colour coordinates, which you want to relate to a single parameter - degree of oxidation. If you collect images for different degrees of oxidation, you will have a system of equations of the type Deg.Ox.(i) = F(R(i), G(i), B(i)), where you want to find the function F. I think something like principle component analysis could be useful for that. Be very careful about the way how the images are acquired, to make sure the lighting condition and camera white balance etc. remain consistent.
Hi, thank you for your answer Radek Machan, and yes, I want to relate the color to the degree of oxidation, so I can compare different colors (like in the image) and thus know which conditions can lead to the least oxidation of the aloe formulations. Browner tones are more oxidized, while the whiter ones are the least. I am going to take the images with a digital camera, so I am quite sure images will be in the RGB format. I didn't think about the possibility to construct a calibration curve, so I will try that strategy, seem reasonable. I have already thought about the conditions to take the images, but thank you for the advice, it is really important.
Hello Ivan Kubovský , thank you very much for your advice. I'll check the characteristics of the spectrophotometer you mentioned. Have a nice day too!
Hi! I don’t know if this is very scientific-like but you can use a free software called Image J to convert an area of your image into a single color (see video attached) and then you can obtain the R, G, B values of that color with Paint (use the tool "Selector de Color", sorry I don't know it's name in English) or any other photo editing software
- Badges
- Science topic
- Similar topics
- Computer Science
- Images