Take quantative measurements of colour from image

Support
1

I am currently working on a project that involves fluorescent dyes emitting in the 400-550 nm wavelength range, specifically green light. As part of this project, I need to image and visualize the fluorescence emitted by the dye and simultaneously measure its optical density.

Considering my research is in its initial stages, I am seeking recommendations for a suitable starter kit that can fulfill the following requirements:

  1. Fluorescence Imaging Capability: The kit should include a fluorescence imaging system capable of visualizing the emitted green light (400-550 nm wavelength) from the fluorescent dye. Ideally, it should offer high sensitivity and resolution for precise imaging.
  2. Optical Density Measurement: The kit should come with a reliable optical density measurement tool or method that allows me to quantitatively measure the concentration of the fluorescent dye accurately.
2 Likes
2

Hi Zikri,

This sounds like an interesting project and perhaps a bit more academic than most maker or hobbyist builds we usual help with, I have some ideas on how you can get the data you need though.

Since the wavelength of light you are trying to measure is well within the limits of human vision, a standard camera should be fine to capture the images. A higher resolution camera will be an advantage to capture more data points but it will likely involve some experimental trial and error to get the distance and camera settings sorted out.

Once you have captured an image of your sample you can extract the colour values for each pixel in the image and store it as an array of data. A quick google search has turned up a few tools designed to do this.

From there you can likely write a program to consider the colours reported for each pixel, and take a measurement of the amount of dye present in the photo that way.

We have a colour sensor that we make in house. One of the challenges for writing the code to recognise the colour correctly was accounting for differences in brightness which would alter the colour data received. Michael discussed how this was mitigated using a HSV format for colour values rather than the standard RGB representation. Iā€™d highly recommend checking out his video below for the full explanation of the challenges.