Identifying Rice Grains of Various Varieties and Studying the Genotype-Phenotype Association of Rice Grains
Date Issued
2016
Date
2016
Author(s)
Kuo, Tzu-Yi
Abstract
Rice (Oryza sativa L.) is a major staple food and is traded globally in considerable amount. Rice shows remarkable variation in grains. The phenotypic information of the rice grains need to be quantified as the first step to investigate the association between the phenotypes and genotypes. This study proposed to distinguish the rice grains of 30 varieties nondestructively using image processing, sparse representation based classification (SRC) and a procedure to phenotype rice grains of 255 varieties in high precision. SRC is a method that uses over-complete bases to capture the representative traits of rice grains. In the experiments, rice seeds were acquired from Genetic Stocks Oryza germplasm collection. The genotypic information (i.e., SNPs) of these seeds are publicly available. The images of the grains were acquired in high resolution using microscopy (approximately 2413 dots per inch). Morphological, color, and textural traits of the grain body, sterile lemmas, and brush were quantified. The traits were subsequently fit into a unified mixed linear model for investigating the association between the phenotypic and genotypic variations of the varieties. An SRC classifier was developed to identify the varieties of the grains using the traits as the inputs. The proposed approach could discriminate the varieties of the rice grains with an accuracy of 89.1%.
Subjects
Variety identification
sparse coding
locality constraint
machine vision
machine learning
image processing
phenotyping
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-105-R02631039-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):ef199fa8cd8219ff58a4cec2c4c25c52