Rice Quality Inspector Using AI/ML & IOT

The Rice Quality Inspector (Grain Insight) project provides an intelligent, automated, and reliable method for assessing the quality of rice grains using a combination of YOLOv8-based image processing and IoT sensor integration. The system monitors essential environmental parameters such as temperature, humidity, and gas concentration through sensors like DHT22, MQ135, and the soil moisture sensor, ensuring that the storage conditions are ideal for maintaining rice quality. Meanwhile, the YOLOv8 model performs accurate grain detection and impurity identification, reducing manual inspection errors and improving consistency in grading and evaluation. Furthermore, the inclusion of AI-driven suggestions for improvement adds a predictive and analytical layer to the system, helping users make better decisions in real time.

[1] “Vision-Based Defect Classification and Weight Estimation of Rice Kernels” — Xiang Wang et al., arXiv (2022). Deep-learning approach that detects kernel defects and estimates defective mass fraction — useful for linking detection to quality metrics.  [2] “Rice Grain Detection and Counting Method Based on an Improved YOLO Model” — Zou et al., Sensors (MDPI), 2023. YOLO-based detection adapted for rice grains (improvements for accuracy/recall) — directly relevant if you use YOLO family models. [3]  Rice Grain Type and Grading of Rice Grains using Image Processing (C. Sindhu   et al., 2021, IJRESM) This paper presents a method to classify and grade rice grains based on their physical features using image processing techniques. [4] Rice Grains Detection, Classification, and Quality Prediction Using Deep Learning (Ramya Asalatha Busi et al., 2024, IJRASET) This study uses deep learning models to automatically detect and classify rice grains and predict their quality. [5] Rice Grain Detection and Counting Method Based on TCLE–YOLO Model — Zou et al., Sensors (MDPI), 2023. This paper proposes an improved YOLOv5 model (TCLE–YOLO) for precise rice grain detection and counting with very high accuracy.