This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
To detect eye (of fish or human or anything else), I think the best approach (with OpenCV, at that time) is to train a Cascade classifier. You need samples, both positive and negative to train your classifier. You could use the Haars features or the LBP with the C++ interface (which you should used!). Almost everything is code for you, you just have to provide data as input, with the label. For the detection, see this page. Enjoy, and let us know if it's working for fish quality!
If it's working, you definitively should make an Android apps, it would be perfect for fish market... ;-)
