I'll outline what you need
those types of recognition are done by machine learning, just some maths using computation. You normally give your program some images to train. Instead of letting the program to figure out by itself, you give it some realistic drawings, the best of the bunch and a picture of the real thing. It will then compare the difference between the real picture and the drawings and get an idea of what good drawings are. If you have some 40 pics, you use 30 to train and some 10 to test. Now when you use the program, when you throw at it a new drawing, it will rank it based on what it learnt. For the raspberry pi part i think it's convenient to move around but not really needed. >From my experience with these kinds of projects, google "image recognition with python" maybe you'll see opencv etc. before that just make sure to learn at least some basic python. Then you google concepts you don't know. little by little, you'll understand what's going on. like take an image recognition tutorial with image set provided and complete it. That approach is better than taking a machine learning tutorial etc. Should you need any help, please tell! I don't promise but i'll try to write such a project as a tutorial. Yours, Abdur-Rahmaan Janhangeer http://www.pythonmembers.club | https://github.com/Abdur-rahmaanJ Mauritius _______________________________________________ Tutor maillist - Tutor@python.org To unsubscribe or change subscription options: https://mail.python.org/mailman/listinfo/tutor
