AUTOMATIC CLASSIFICATION OF PLANT-CELLS ACCORDING TO TISSUE-TYPE USING ANATOMICAL FEATURES OBTAINED BY THE DISTANCE TRANSFORM

Image analysis was used to develop a faster and more objective method for the quantitative measurement of plant anatomy. The size, number an d anatomical features of individual cells were measured automatically using methods based on image skeletonization and the distance transfor m. The variation in cell wall thickness around individual cells was me asured by masking the distance transform of a segmented binary cell wa ll image with a skeleton of the same image to extract distance values along the boundary between neighbouring cells. The measurements were u sed to create a multi-dimensional feature space in which individual ce lls were classified automatically according to their tissue type. Cell s were classified into seven types using discriminant analysis, and th e performance of the classification rule was examined by cross-validat ion. The potential use of these methods as a research tool, and in pla nt breeding programmes is discussed. (C) 1996 Annals of Botany Company .