Pedestrian Detection System with a Clear Approach on Raspberry Pi 3

Object detection (OD) is an important application in many fields of study. An implementation of OD follows in different methods. Histogram of gradients (HOG) and machine learning algorithms shows us the prominent features in object detection. In this paper the implementation of the HOG with machine learning techniques is explained on MATLAB and OpenCV framework on Raspberry Pi 3 board (RPI3). An application of pedestrian detection (PD) is implemented with the detection of humans from the video. In training stage, HOG extracts the features from the images, then trained on Support Vector Machine(SVM) with those features. In detecting stage, video to frame, sliding window, non-max suppression, HOG and SVM analysis are executed.