SVRice: An Automated Rice Seed Vigor Classification System via Radicle Emergence Testing Using Image-Processing, Curve-Fitting, and Clustering Methods

Highlights SVRice package: a simple and effective procedure for automatic rice seed vigor classification was verified. The package consists of three important parts: image analysis, curve-fitting, cluster analysis. Findings provide an system that promises to reduce the production costs. Abstract. Rice is storage management by producers farmers while planning their cultivation activities. Field emergence direct method testing but laborious, time-consuming subjective. This article presents package, simple, cost-efficient, flexible based on computer analysis high-throughput, classification. four steps: dynamic imaging, processing, curve fitting, clustering. Seed classified radicle indices, such as maximum (MaxRE), mean time (MRET), speed (t 50 ), uniformity (U 7525 area under fitted (AUC). Parameters used classify vigor, MRET, U , t were strongly negatively correlated with saturated salt accelerated aging (SSAA) test. A germination 90 h at 25°C sufficient SVRice, whereas SSAA test took approximately 400 complete. software algorithm created be especially suitable assessment after six months controlled atmosphere (at 15°C 37% RH in hermetic bag). study showed could unambiguously 40 indica samples different varieties, years, sites, times, conditions compared costs encourages competitiveness producers. Keywords: Image Oryza sativa, Radicle emergence, SVRice.