Hello! I work in medical image processing and machine learning field, grouped together. So, getting to the task:
Regarding your project:
1) We do bit of preprocessing on the images :
a) scale to higher pyramid with stronger features
b) apply cosine transform
c) consider low frequency components only
2) We use the features designed below, instead of directly feeding the images:
a) Run length encoded vector (nx2 dim)
b) spatial location and edge length (obtained from edge map)
3) Train SVM classifier
This is the basis of the work. I shall made changes and modification when I take a look at some sample images/data (Thank you for the attachment, please attach some more zip them in case needed).
Queries:
When is the deadline? What documentation you would need? Do you need any GUI wrapped over it? All these shall affect the budget and timeline.
Tools:
MATLAB/Python
NOTE:
As a principle, I first give a timeline and stick to it regarding progress, update and review and milestone payments. Given the nature of work you should be able to communicate timely.
So, if you're interested, kindly leave your queries and clarifications, if I'm not online still, and I shall get back ASAP.
Let's get some work done :)
Thank you! Have a great weekend!