Optical signal intensity incorporated rice seed cultivar classification using optical coherence tomography

Abstract

Here, the optical signal intensity of swept-source optical coherence tomography (SS-OCT) was incorporated to assess the images of different rice seeds for precise cultivar classification. Such classification is essential from various agricultural viewpoints but remains challenging because rice seeds have a similar visual appearance. Thus, studies on rice seed cultivar classification play a key role in achieving a successful rice yield. To investigate the feasibility of the proposed approach, two-dimensional OCT images were acquired from two types of rice seed cultivars (Chucheong and Hopyeong), which were then subjected to an optical signal intensity assessment. Based on the depth-dependent optical scan profile results, the averaged and normalized back-scattered light intensity curves revealed a distinct difference between each cultivar, which aided the precise classification of the cultivars. Therefore, the developed optical signal-sensing approach can be extensively applicable to seed cultivar classification in the agricultural industry.