People Detection with DSIFT Algorithm

People detection is an interesting computer vision topic. Locating people in images and videos have many potential applications, such as human computer interaction and auto-focus cameras. There have been much effort on developing people detection algorithms, and among them, the “Dense ScaleInvariant Feature Transform (DSIFT) [1]” algorithms appear to be very effective. To implement our own people detection program, we utilized the DSIFT algorithm to extract feature vectors from images, and used Support Vector Machine (SVM) and Naive Bayes algorithms for classification. We explored different imaging processing and machine learning techniques, and assessed their performance.