%0 Journal Article
%T A novel machine-vision-based facility for the automatic evaluation of yield-related traits in rice
%A Lingfeng Duan
%A Wanneng Yang
%A Chenglong Huang
%A Qian Liu
%J Plant Methods
%D 2011
%I BioMed Central
%R 10.1186/1746-4811-7-44
%X Rice is the staple food for a large number of countries and regions in the world, particularly in Asia [1]. Because the world's population is increasing, obtaining higher yields has been the primary breeding target of rice cultivation [2]. As a complex agronomic trait, rice yield is determined by the product of the grain weight, the number of grains per panicle and the number of panicles per plant. The number of total spikelets per panicle and the seed setting rate are two traits that multiplicatively determine the number of grains per panicle, and the grain weight is largely determined by the grain size, including the grain length, the grain width, and the grain thickness [3].The evaluation of yield traits, including the number of total spikelets (including filled and unfilled spikelets), the number of grains (also known as the number of filled spikelets), the seed setting rate (the number of filled spikelets divided by the number of total spikelets), the 1000-grain weight, the grain length, and the grain width, is an essential step in rice breeding, genetic research and functional genomics research [4-6]. Currently, rice yield trait evaluation is mainly performed by experienced workers. When investigations of large numbers of plants are needed, the manual measurement process is very subjective, inefficient, tedious, and error-prone. Most importantly, manual measurements are greatly affected by worker fatigue, which is a major problem in conducting mass measurements and renders the evaluation results questionable. In addition to trait extraction and evaluation, data logging and seed management are two instrumental steps in rice research. Traditionally, the processing of data, seed packaging, and seed coding are preformed manually and are thus error-prone and unreliable. A mistake in data management and seed management would lead to incorrect decisions and treatment of the seeds and is thus intolerable in rice research. For this reason, at least three workers are no
%K Rice
%K Yield-related traits
%K Fast trait evaluation
%K Plant phenotyping
%K Machine vision
%U http://www.plantmethods.com/content/7/1/44