Department of Image Processing

Recognition of partially occluded and deformed binary objects

Journal Article

Horáček Ondřej, Kamenický Jan, Flusser Jan

serial: Pattern Recognition Letters vol.29, 3 (2008), p. 360-369

research: CEZ:AV0Z10750506

project(s): 1M0572, GA MŠk

keywords: Image recognition, Partial occlusion, Affine transformation, Inflection point, Radial vector

preview: Download

abstract (eng):

A method dealing with recognition of partially occluded and affine transformed binary objects is presented. The method is designed for objects with smooth curved boundary. It divides an object into affine-invariant parts and uses modified radial vector for the description of parts. Object recognition is performed via string matching in the space of radial vectors.

abstract (cze):

Rozpoznávání částečně zakrytých a deformovaných binárních objektů

RIV: IN