A non-rigid registration method for multispectral imaging of plants
Autor: | Jonas De Vylder, Dominique Van Der Straeten, Filip Vandenbussche, Koen Douterloigne, Wilfried Philips |
---|---|
Přispěvatelé: | Kim, MS, Tu, SI, Chao, K |
Rok vydání: | 2012 |
Předmět: |
business.industry
Computer science Multispectral image ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Biology and Life Sciences Image registration non-affine registration Set (abstract data type) image registration Transformation (function) Feature (computer vision) SPLINES ARABIDOPSIS-THALIANA multispectral imaging GROWTH Computer vision Artificial intelligence business |
Zdroj: | Proceedings of SPIE, the International Society for Optical Engineering |
ISSN: | 0277-786X |
Popis: | Registration of multispectral images remains a challenging task due to the lack of stable feature points. Methods based on intensities are generally more robust for multi-modal image registration, but are computationally demanding or are restrictive to the transformation model allowed in the registration. This paper proposes a new registration framework which overcomes these drawbacks. The proposed method optimizes the location of a set of virtual landmarks in order to get robust and accurate registration. |
Databáze: | OpenAIRE |
Externí odkaz: |