Implementation of Dalal and Triggs Algorithm to Detect and Track Human and Non-Human Classifications by Using Histogram-Oriented Gradient Approach

Abstract

In this paper, the proposed application is mainly used for security purpose and it deals with searching based on the given input and integrates this approach with the concept of histogram of oriented gradient (HOG) features to achieve person or non-person classification in human tracking system. It captures the features of human automatically based on the gradients of an image in human tracking system. This tracking system process will be done only through the expert team. Before getting into the process, the images or videos which have been captured from the surveillance camera have to be uploaded and checked whether it is a human being or not. Two leading approaches to the application are: Single-detection window approachPart-based approach. This application process is done based on some scenarios like color code, distance between the object and the camera, and the height of the human being.