Development of an Integrated Image Acquisition Setup for Assigning Selling Price of Rough and Milled Rice (Oryza Sativa L.) in the Supply Chain

In agro-processing industries, cereal grains acquire a significant place but, many times the consistency in grain quality gets compromised during the supply chain, owing to improper and/or inefficient supervision. In the current work, an integrated computer vision setup, aided with the gravimetric principle, has been developed for assigning the best price for rough and milled rice in the supply chain hierarchy. Three local paddy varieties were classified based on physical and color features extraction, following established image processing techniques. Two unique features i.e., dry mass identifier (DMI) and head rice equivalent (HRE) were introduced for assessing inter-varietal paddy admixture, grain moisture content, and broken fractions at several size segments to achieve consistent pricing. Both features have shown a good correlation with the manual classification which is although accurate but lengthy and tedious. The devised setup can be adopted as an inexpensive, non-invasive, and non-destructive tool in rice trading.