I am new in Machine Learning. I have taken a course in vision and we are required to do a project.
I am thinking of data mining medical lab report images. My code must take an image and jpg file and then extract important information from it like lab where test has been done, patient name, test type and more important various data like heamoglobin, RBC, etc in case of blood test report.
I can build an OCR, but, problem which I am stuck at is in case of data which generally forms a table like structure. So, I want to find that tabular structure on which I can just apply matrix extraction to find various datas.
I'm looking for assistance with two basic things:
Is my approach of finding tables and then extracting data is correct? If yes, then can you point out some good papers or implementation to find tabular structure. (P.S.- Don't mention tabular)
Any approach which is state-of-the-art or good? (Paper or implementation)