Profile cover photo
Profile photo
Roman Trigubenko

Post has attachment
Hello everyone! Here is future Ph.D at Computer Science. To to be honest, i will start it only from the next year, but now trying to create interesting projects.
By the way, i am starting new one and need in few tip`s and advices from people, who already good in it :3 I already created (not from 0, shame on me) face detection very simple framework and now have new target - create from 0 program to check condition of something by single image! For example i have list of paper. When its pure white, rectange, without any defects and paint on it - that should be a standard. If i will check another photo with another list of paper, same size and same position, but with paint on it, or torn in the middle, program should write, that Dude, we have a problem here. Kind of check, as i said.
I wrote it because want to create project from 0! Really exiting and full of power! I will try to use tensorflow, openCV. But, i dont know how to realise this check mechanic. For example, if match percent (standard -> current list of paper) is <=70%, that has defect or kind of. I will be happy to hear hints, advices at writing/train model, realisation of it! I am just newby, but studying hard.
Thank you very much for reading it and, for sure, have a great day! :)
From Japan, with love <3

Post has attachment
Wait while more posts are being loaded