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

Chongyuan Zhang; Sindhuja Sankaran

doi:10.1007/978-1-0716-2537-8_8

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

Methods Mol Biol. 2022:2539:71-76. doi: 10.1007/978-1-0716-2537-8_8.

Authors

Chongyuan Zhang¹, Sindhuja Sankaran²

Affiliations

¹ Department of Biological Systems Engineering, Washington State University, Pullman, WA, USA.
² Department of Biological Systems Engineering, Washington State University, Pullman, WA, USA. sindhuja.sankaran@wsu.edu.

PMID: 35895197
DOI: 10.1007/978-1-0716-2537-8_8

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.

Keywords: Feature extraction; Image processing, Grains; Seed phenotyping.

MeSH terms

Crops, Agricultural
Humans
Image Processing, Computer-Assisted / methods
Phenotype
Plant Breeding*
Seeds*