I need to quantify the browning of an apple

Hi, I am new to ImageJ and am doing an experiment to decrease the level of browning in apple slices. I take before and after pictures of apple slices and need to calculate the level of browning in the slices at the end. The only problem I have is the pictures are not the same brightness every time even though the camera is in the same spot with controlled lighting, so when I measure the means of each individual slice before and after the difference is not as drastic as hoped even when the apple has browned significantly.


Is there any way to adjust the images to compensate for the different light intensities? Thanks for any help.

Hello,
I think your problem is related to the conditions the images are taken. The illumination is not even, and even this is not the same for both images. The camera may also be in cause (parameters of the white balance).
If I were you, I would to take new pictures, one for each apple at full resolution and closer distance. The illumination variations would probably be less problematic. It’s better to take good pictures than trying to process them to make them good.

If you cannot reproduce the images, you can at least take a picture of the background without apples, with exactly the same settings, then try to find a way to use it to correct the images (not sure how I would do that with color images, though).

But, if you really need to modify the images digitally now they’ve been taken, you can try Process>Subtract Background. Again, not sure how reliable the results will be.

One last point: If I zoom on the apples, they all don’t have the same texture, i.e. some of them are a little grainier/shinier than the others, possibly due to droplets of juice exsuding from the slice and the rugosity of the surface. I don’t know if it’s important, but be aware of that. If you take new pictures with one apple at a time, try to swipe it and check if it makes a difference. Also, be careful when quantifying the brown content: Don’t include the pips !

Useful tools : Analyze>Color histogram, Plugins>Color Inspector

Finally, if possible, don’t save images in JPG format (lossy format).

Hello,
As Nicolas wrote, getting the best possible photos will simplify the analysis process later on. While I can’t help you on ImageJ, I can give you some advice on taking the pics.

  1. Shoot in a controlled light environment (always the same lamp intensity, shielded from outside light)
  2. Shoot in Manual mode
  3. Shoot in RAW mode rather than JPEG

[3] is the most important. In JPEG mode, the camera will make multiple “intelligent” adjustments, which will modify the whole picture, even in a stable light environment, just due to the very browning modifications you’re trying to track.

Then process your RAW pictures with a tool like Adobe Lightroom, applying the same develop parameters to every picture, and you’ll get easily comparable pictures (Lightroom allows export to TIFF or JPEG for further processing in ImageJ).

Hope this helps,
Phil