An Online Intelligent Rice Classification System
Rice is a major food source worldwide. With an increased production of rice, international rice market faces challenges in quality control due to mixing of various available rice varieties. Manual rice classification methods are a solution to one end, but they are time consuming, prone to error and lack scalability. This project tackles this issue with an AI powered solution using deep learning. The process involves data collection, data preprocessing, model training and testing, and web app integration. This user-friendly web platform uses two deep convolutional neural network CNN models i.e. ResNet50 and YOLOv8 to automate rice classification process. The users can upload scanned rice images and get the accurate classification results. This project majorly focuses on popular Basmati varieties in Asia like Basmati 2000, Kainat 1121, Sella 1509, and Sella Basmati 1121, with their distinct features. The project addresses challenges in verifying rice varieties, reducing operational costs, and modernizing agriculture. Potentially, it serves as a valuable tool for farmers, consumers, exporters, traders, and rice mill owners, ensuring food quality and safety through accurate rice variety identification.
Keywords: Deep Learning,Convolutional Neural Network,Image Processing, Streamlit Web Application,ResNet50, YOLOv8,Rice Classification, Basmati Rice
Tools: Python,Tensorflow/Keras,Streamlit,Yolov8,ResNet50,Flatbed Scanner
Department: Department of Electrical Engineering
Project Team Members
| Name |
Email |
|
Kamran Younis
|
kamran2020@namal.edu.pk |
|
Usman Shafique
|
usman2020@namal.edu.pk |
Project Poster
