Hi
I have 4 years experience in Image Processing. i've worked on similar project before.
Libraries: Opencv, Tensorflow, Dlib etc
Languages: C++, Python, C#, C++/Cli
Framework: Qt, .Net
Out of all traditional algorithms (HOG, Haar, Bag of Visual words etc) i prefer more advanced models from convolution neural network.
With A.I models we can get better accuracy, how much dataset of disease do you have?
I've written several models in CNN, i can write custom model which works best with your hardware (speed vs accuracy trade off)
Let me know in chat.
Thanks & Regards
Prakash