Object detection and identification using SURF and BoW model

Object detection and identification is a fundamental workflow in Computer vision. In this paper I am presenting a feature based approach to detect an object in cluttered scene using “Speeded Up Robust Features (SURF) and to identify object in real time manner using Bag-of-words (BoW) model. The System trains the model with different supervised Machine learning classifiers like Support Vector Machine (SVM) and k-nearest neighbors and compares their performance. I used Computer Vision and Machine Learning toolboxes of Matlab (2015a).