Year: 2025
Type: article
Abstract: Rice quality grading is a critical step in ensuring the market value and consumer acceptance of rice. It involves assessing various physical attributes such as grain integrity, polish level and shape to determine its suitability for consumption, processing, or export. Traditional methods of identifying unpolished and broken rice grains rely on manual inspection, which is time-consuming and prone to human error. This article suggests an automatic segmentation of unpolished and broken rice grains using contour-based pixel distribution approach. The method leverages image processing techniques such as greyscale conversion, adaptive thresholding and morphological operations to extract individual grains from a sample. Contours are detected using the Canny edge detection algorithm, followed by connected component analysis to segment grains based on their shape and size. To differentiate the unpolished grains, pixel intensity distributions are analysed, as unpolished rice exhibits higher roughness and uneven textures. Similarly, broken grains are identified based on contour perimeter, aspect ratio and area thresholds. The proposed system was tested on rice samples collected from the VIT School of Agricultural Innovations and Advanced Learning (VAIAL), Vellore and NK Rice Mill, Redhills, Chennai, achieving high accuracy in distinguishing whole, broken and unpolished grains. This approach provides a cost-effective, non-destructive and efficient solution for automated rice quality grading, reducing dependency on manual inspection.
Source: Plant Science Today
DOI: 10.14719/pst.7844
Institution: Vellore Institute of Technology University
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: diamond
OpenAlex ID: W4416564705