Near-Infrared Hyperspectral Imaging Pipelines for Pasture Seed Quality Evaluation: An Overview

Abstract

Near-infrared (800–2500 nm; NIR) spectroscopy coupled to hyperspectral imaging (NIR-HSI) has greatly enhanced its capability and thus widened its application and use across various industries. This non-destructive technique that is sensitive to both physical and chemical attributes of virtually any material can be used for both qualitative and quantitative analyses. This review describes the advancement of NIR to NIR-HSI in agricultural applications with a focus on seed quality features for agronomically important seeds. NIR-HSI seed phenotyping, describing sample sizes used for building high-accuracy calibration and prediction models for full or selected wavelengths of the NIR region, is explored. The molecular interpretation of absorbance bands in the NIR region is difficult; hence, this review offers important NIR absorbance band assignments that have been reported in literature. Opportunities for NIR-HSI seed phenotyping in forage grass seed are described and a step-by-step data-acquisition and analysis pipeline for the determination of seed quality in perennial ryegrass seeds is also presented.

Single Seed Near-Infrared Hyperspectral Imaging for Classification of Perennial Ryegrass Seed

“…Cultivar identification performed in other seed applications has also shown 80–100% levels of calibration and prediction accuracy [ 6 ]. A study that discriminated between rice seed cultivars of the same age to avoid batch effects using NIR-HSI covering wavelengths 874–1734 nm resulted in 80% classification accuracy using PLS-DA and KNN and 100% accuracy using Soft Independent Modeling of Class Analogy (SIMCA), RF and SVM classifiers.…”

“…In pasture grasses, such as ryegrass, varietal purity extends to endophyte content and purity; thus, batch purity of high-quality cultivars is important in marketed forage species. While there are established quality control measures for cultivar discrimination [ 4 ] and endophyte content and purity [ 5 ], using sensitive tests such as DNA-based genotyping is often complex, labour intensive, expensive and destructive, thus an evaluation is limited to a randomly selected subset of seeds from a batch and not all batches are tested [ 6 ]. These attributes are not ideal when seeds are valuable and/or the supply is limited.…”

“…With the advantages of being a non-destructive and potentially high-throughput technique for the acquisition of both qualitative and quantitative data, NIR-HSI investigations for seed quality assessments have become increasingly popular. As reviewed by Reddy et al [ 6 ], NIR-HSI has been applied in the varietal classification of plant species, such as cotton and maize, as well as the detection of disease-causing fungal pathogens on seed [ 10 , 11 , 12 , 13 , 14 , 15 , 16 ]. However, the potential of the technology for use in seed-based cultivar discrimination in pastures, such as perennial ryegrass and, importantly, beneficial endophyte detection, has not been explored.…”

“…There is good evidence in the literature that varietal identification can be achieved with at least 80% accuracy [ 6 ]. Zhu et al [ 17 ] identified seven varieties of cotton seeds using NIR-HSI wavelengths 942–1646 nm.…”

EXNet: (2+1)D Extreme Xception Net for Hyperspectral Image Classification

“…Given such advantages, HSI has extensive application across various real-world domains. This includes environmental studies [3], [4], atmospheric research [5], [6], agriculture [7], forestry [8], assessments of water quality and contamination [9], mineral exploration, food processing and quality assessment [10]- [17], medical applications [18]- [21], and even military domains [22], [23].…”