Автор: Gutierrez JosГ© A., Armstrong Brian S.R.
Название: Precision Landmark Location for Machine Vision and Photogrammetry / Finding and Achieving the Maximum Possible Accuracy
ISBN: 1846289122 ISBN-13(EAN): 9781846289125
Цена: 15372 р.
Наличие на складе: Есть у поставщика Поставка под заказ.
Описание: The applications of image-based measurement are many and various: image-guided surgery, mobile-robot navigation, component alignment, part inspection and photogrammetry, among others. In all these applications, landmarks are detected and located in images, and measurements made from those locations.Precision Landmark Location for Machine Vision and Photogrammetry addresses the ubiquitous problem of measurement error associated with determining the location of landmarks in images. With a detailed model of the image formation process and landmark location estimation, the CramГ©rвЂ“Rao Lower Bound (CRLB) theory of statistics is applied to determine the least possible measurement uncertainty in a given situation.This monograph provides the reader with:вЂў the most complete treatment to date of precision landmark location and the engineering aspects of image capture and processing;вЂў detailed theoretical treatment of the CRLB;вЂў a software tool for analyzing the potential performance-specific camera/lens/algorithm configurations;вЂў two novel algorithms which achieve precision very close to the CRLB;вЂў an experimental method for determining the accuracy of landmark location;вЂў downloadable MATLABВ® package to assist the reader with applying theoretically-derived results to practical engineering configurations.All of this adds up to a treatment that is at once theoretically sound and eminently practical.Precision Landmark Location for Machine Vision and Photogrammetry will be of great interest to computer scientists and engineers working with and/or studying image processing and measurement. It includes cutting-edge theoretical developments and practical tools so it will appeal to research investigators and system designers.