Abstract
Geometric moments have been proven to be a very efficient tool for description and recognition of binary shapes. Numerous methods for effective calculation of image moments have been presented up to now. Recently, Sossa, Yañez and Díaz [Pattern Recognition, 34(2):271-276, 2001] proposed a new algorithm based on a morphologic decomposition of the image into a set of closed disks. Their algorithm yields approximative results. In this paper we propose a refinement of their method that performs as fast as the original one but gives exact results. © Springer-Verlag 2004.
Cite
CITATION STYLE
Sossa, H., & Flusser, J. (2004). Refined method for the fast and exact computation of moment invariants. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3287, 487–494. https://doi.org/10.1007/978-3-540-30463-0_61
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
