High-Throughput Extraction of Seed Traits Using Image Acquisition and Analysis.

Abstract

Seed traits can easily be assessed using image processing tools to evaluate differences in crop variety performances in response to environment and stress. In this chapter, we describe a protocol to measure seed traits that can be applied to crops with small grains, including legume grains with little modification. The imaging processing tool can be applied to process a batch of images without human intervention. The method allows evaluation of geometric and color features, and currently extracts 11 seed traits that include number of seeds, seed area, major axis, minor axis, eccentricity, and mean and standard deviation of reflectance in red, green, and blue channels from seed images. Protocols or methods, including the one described in this chapter, facilitate phenotyping seed traits in a high-throughput and automated manner, which can be applied in plant breeding programs and food processing industry to evaluate seed quality.

Plant phenotyping relevance

種子画像から形状・色など11形質を自動抽出する高スループット画像解析プロトコルが中心であり、植物フェノタイピング手法に該当する。

Code and data availability

公開本文の所在を確認できませんでした。非公開または購読が必要な可能性があります。

No evidence-backed public reproduction asset is currently recorded.