I want to use either OpenGL, OpenCV, or OpenMAX for just reading a jpeg image, and convert it to RGB raw data.
I know that OpenCV is more of a computer vision library and the other ones are more of a rendering tool, but I believe that they all can be used if the purpose is just for decoding the image, is this right?
Please feel free to correct me.
Now, if the purpose is just pure decoding an nothing more, then how do I know which library will perform relatively faster than the other ones?
All I care about is to have a faster decoder. Can someone tell me which one should I use?
Also, now that we have both CPU and GPU, how can I make use of the availability of GPU, and make sure that decoding will be handed over to GPU for faster processing?
Is there a certain library that I can use to make sure that the decoding will be done in GPU?
Is GPU only helpful, if we are trying to display images? Or it could be equally helpful even if we are only doing decoding of jpeg images, without displaying them?
Basically, how to utilize GPU to its fullest when it comes to decoding images? And, which library will be the fastest? How about libjpeg library?
I would appreciate some help.