Robotics Research Technical Report: Identification of Partially Obscured Objects in Two Dimensions by Matching of Noisy 'Characteristic Curves' (Classic Reprint) - Softcover

J. T. Schwartz

 
9781332142248: Robotics Research Technical Report: Identification of Partially Obscured Objects in Two Dimensions by Matching of Noisy 'Characteristic Curves' (Classic Reprint)

Synopsis

Identify partially obscured objects in two dimensions by matching noisy curves to models, with fast, robust algorithms that handle occlusion.

This guide presents a boundary- and curve-based approach to recognizing objects when parts are hidden from view. It focuses on using rigidly embedded curves on object surfaces that move with rotation and translation, enabling fast, accurate matching even in noisy images. The method emphasizes practical smoothing, least-squares matching, and efficient computation suitable for real-time analysis.

Key ideas include working with 2-D and 3-D scenarios, defining curves along object boundaries, and using smoothing to counteract noise. The approach leverages fast Fourier transforms to reduce matching time and discusses how to extend concepts to compound scenes where objects overlap or are partially obscured.

  • Understand how characteristic curves are chosen and made robust to noise.
  • Learn how to parametrize curves, compute optimal alignment, and handle partial visibility.
  • Explore smoothing techniques that recover true curves from noisy observations.
  • See how the method scales with model complexity and supports both 2-D and 3-D cases.

Ideal for readers applying computer vision and pattern-minding techniques to recognize objects in cluttered or partially occluded scenes.

"synopsis" may belong to another edition of this title.

Other Popular Editions of the Same Title

9780484520935: Robotics Research Technical Report: Identification of Partially Obscured Objects in Two Dimensions by Matching of Noisy 'Characteristic Curves' (Classic Reprint)

Featured Edition

ISBN 10:  0484520938 ISBN 13:  9780484520935
Publisher: Forgotten Books, 2017
Hardcover