Extract reliable points from noisy curves to boost 3D object recognition.
This edition explains how to represent a group of curves with a small set of points that stay meaningful despite noise. It compares two practical methods for picking these representative points and tests them with synthetic data to show how they perform under different noise levels. The goal is to speed up matching against models while keeping accuracy in three-space.
- Learn two concrete methods for selecting representative points: a centroid-like approach using nearby curve points and a method that follows closest-approach points along smooth curve fits.
- See how noise and how much of the curve is used affect matching quality and the stability of the chosen points.
- Understand how a separation measure between two curve sets helps evaluate how well observed data matches a model.
- Discover how synthetic data is generated to test robustness and what that implies for real-world sensing and robotics applications.
Ideal for readers working in robotics, computer vision, or object recognition who want practical techniques for robust feature extraction from noisy 3D curves.
Les informations fournies dans la section « Synopsis » peuvent faire référence à une autre édition de ce titre.
Vendeur : Forgotten Books, London, Royaume-Uni
Paperback. Etat : New. Print on Demand. This book explores a novel approach to object recognition by describing objects in terms of their 3D curves. It presents two methods for choosing representative points of closest approach that can be used to efficiently match sets of curves in 3D space, even when the curves are corrupted by noise. The methods are evaluated using computer-generated curves with varying amounts of noise, and the results demonstrate that the centroid method allows better selection of points than quadratic or cubic fits when substantial lengths of the curves can be used, but that a cubic fit of coordinates vs arc length gave better results when relatively short lengths of curve were used. The quadratic fits behaved very badly. The book provides a valuable contribution to the field of object recognition and has applications in data reduction, efficient recognition of 3D objects, and other areas where measuring the spatial separation of sets of curves in 3D space is important. This book is a reproduction of an important historical work, digitally reconstructed using state-of-the-art technology to preserve the original format. In rare cases, an imperfection in the original, such as a blemish or missing page, may be replicated in the book. print-on-demand item. N° de réf. du vendeur 9781334537950_0
Quantité disponible : Plus de 20 disponibles
Vendeur : PBShop.store US, Wood Dale, IL, Etats-Unis
PAP. Etat : New. New Book. Shipped from UK. Established seller since 2000. N° de réf. du vendeur LW-9781334537950
Quantité disponible : 15 disponible(s)
Vendeur : PBShop.store UK, Fairford, GLOS, Royaume-Uni
PAP. Etat : New. New Book. Shipped from UK. Established seller since 2000. N° de réf. du vendeur LW-9781334537950
Quantité disponible : 15 disponible(s)