Millimeter wave radar with machine intelligence for home surveillance applications

Abstract

With the fast-developing technology, the application using emerging technology is replacing some application using conventional one. Radar for surveillance application can be one of them. With the help of machine learning and big data, it can reach a very high recognition and classification rate. This project is aimed to develop a relatively low-cost radar system with machine learning and mainly focus on improving the recognition accuracy. The report summarizes the knowledge of 24GHz radar working principle, signal processing and image processing, compares the recognition accuracy result with different machine learning result. As a result, the radar system can reach up to 98% recognition accuracy with limited training data.