:: AForge.NET Framework :: Articles :: Forums ::

Adjust Image Colorimetry from Image Reference

The forum is to discuss topics related to different areas of image processing and computer vision.

Adjust Image Colorimetry from Image Reference

Postby Vogel » Wed Aug 14, 2013 3:03 pm

First sorry for my poor english ;)

I warn you that I do not yet know much about "image processing" apart some information I could find on the internet. My project is to obtain homogeneous Tile composed from many Aerial imagery pictures, the problem is those images are shooted in many different time & weather condition so the colorimetry of many image could be very different between each other. So I need your lights on the following problem:

I would adjust the colorimetry of a group of image using as reference the colorimetry of a single image.


At that time, I can get something decent on the part of "dominant color" of the image but not on contrast and luminosity...

In summary, I have attempted a calculation approach of histograms offset between three RGB channels. In detail, for the Reference image(Ref), I look first for the "mean" value of different channels (R, G, B). Then I calculate the average differences between the Red and green (example: +23) and the green and blue (example: -8). From thoses, I get my Reference Offset Histogram Value.
Then, for each image (X) which I want to apply my reference colorimetry, I calculate the same way Offset Histogram Value. And finaly, with (X Offset Histogram Value - Reference Offset Histogram Value) on each channel I get the offset to apply on each pixel from each channel...

My questions:
1- There is another way more quick or more simple or more effecient to do this ?

2- Is Aforge can afford to do this type for treatment?

3- In your opinion is it possible to adjust the contrast and brightness of the image in the same way and how ? And it's a good idea to do it ?

4- Are there other methods of adjusting an image relative to a reference that you seem more effective?

I search from many month on this problem :geek: so any help would be very usefull... thanks guys

Posts: 1
Joined: Wed Aug 14, 2013 12:04 pm

Return to Image Processing and Computer Vision