I really like the idea behind OpenCV, but i am searching for some good object detection tutorials for the Android platform using OpenCV 18.104.22.168. Basically i want to distinguish between a pre- defined set of known objects. I want to train the matcher using the Camera and label the trained objects with some text. I also want to save the trained knowledge base.
After the training phase, i want to use the camera to do some augmented reality stuff. Hence the matching should at least be scale invariant, the best would be also rotation invariant.
I tried to study the documentation the last few days, but i am really confused due to the inavailability of some algorithms on the Android platform and also the very confusing error messages when trying to use the drawMatches() method from the native backend for instance.
Is there any help, example code or good tutorial available for this topic?
Thanks in advance Mirko