Machine Learning: developing an image recognition program : with Python, Scikit Learn and OpenCV

Abstract

Machine Learning is one of the most debated topic in computer world these days, especially after the first Computer Go program has beaten human Go world champion. Among endless application of Machine Learning, image recognition, which problem is processing enormous amount of data from dynamic input.

This thesis will present the basic concept of Machine Learning, Machine Learning algorithms, Python programming language and Scikit Learn – a simple and efficient tool for data analysis in Python. To demonstrate all of these concepts, a small program has been created whose function was to first process a dataset which contain a number of face images, then feed it through a machine learning algorithm and finally using live detection from a webcam to give the live input for the program to predict weather or not the person is smiling.