Low-Cost Machine Vision System for Sorting Green Lentils (Lens Culinaris) Based on Pneumatic Ejection and Deep Learning

TL;DR

This paper introduces a low-cost, real-time machine vision system using deep learning for sorting green lentils into multiple categories with high accuracy, integrating pneumatic ejection and modular hardware for efficient grain separation.

Contribution

It presents a novel two-stage YOLOv8-based classification pipeline combined with pneumatic ejection on a low-cost platform for effective lentil sorting.

Findings

Achieved 87.2% accuracy in grain separation.

Operates at 59 mm/s conveyor speed.

Demonstrates potential for cost-effective grain sorting systems.

Abstract

This paper presents the design, development, and evaluation of a dynamic grain classification system for green lentils (Lens Culinaris), which leverages computer vision and pneumatic ejection. The system integrates a YOLOv8-based detection model that identifies and locates grains on a conveyor belt, together with a second YOLOv8-based classification model that categorises grains into six classes: Good, Yellow, Broken, Peeled, Dotted, and Reject. This two-stage YOLOv8 pipeline enables accurate, real-time, multi-class categorisation of lentils, implemented on a low-cost, modular hardware platform. The pneumatic ejection mechanism separates defective grains, while an Arduino-based control system coordinates real-time interaction between the vision system and mechanical components. The system operates effectively at a conveyor speed of 59 mm/s, achieving a grain separation accuracy of 87.2%. Despite a limited processing rate of 8 grams per minute, the prototype demonstrates the potential of machine vision for grain sorting and provides a modular foundation for future enhancements.