Multiple Target Tracker and Human Classifier for Radar Application
Author | : Jonathan J. Preussner |
Publisher | : |
Total Pages | : |
Release | : 2005 |
ISBN-10 | : OCLC:436098124 |
ISBN-13 | : |
Rating | : 4/5 ( Downloads) |
Download or read book Multiple Target Tracker and Human Classifier for Radar Application written by Jonathan J. Preussner and published by . This book was released on 2005 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: ABSTRACT: Human classification is a valuable tactical tool for military and security applications. Multiple target tracking is inherent to target classification in the radar system. A target tracker is designed in Matlab. It uses a simplified Kalman filter to predict range and velocities for multiple targets. Multiple target tracking allows several targets to be tracked and classified simultaneously. The thesis research addresses the design and implementation of human classifiers. The classification algorithms all use Spectrogram data to classify targets. The Fourier transform method classifies human targets by the periodic motions of human gait. Emphasis will be placed on optimizing the algorithm for implementation in a digital signal processor. Detemining targets' angles of incidence is also discussed. Additional improvements in the application of this classifier are made to reduce the required computational and storage resources. A template matching algorithm is also researched. The template matching method uses previously recorded sets of data to classify targets. This algorithm will be designed for near-real time classification of human and non-human targets. All classification methods require multiple-target tracking to generate Spectrograms for classification.