Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{191641,
        author = {Ravi Ramhari Bhong and Abhishek Narayan Warwade and Satyam shivkumar biradar and Anil mahavir kale and Prof. Indraneel Mane},
        title = {AI/ML-Based Quality Assurance System for Agricultural Products},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {8},
        pages = {9043-9053},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=191641},
        abstract = {Agricultural product satisfactory evaluation is a crucial aspect of supply chain management, marketplace standardization, and customer pride. In India, onions are one of the most economically giant vegetable vegetation, widely cultivated and consumed daily. However, grading and sorting of onions are predominantly performed manually in agricultural markets, storage facilities, and distribution centers. Manual evaluation is extraordinarily subjective and motivated through factors inclusive of inspector revel in, time strain, fatigue, and inconsistent lighting situations. As a end result, the satisfactory classification of onions varies substantially across markets, leading to disputes among traders and suppliers, compromised pricing fairness, and reduced export reliability. To deal with those limitations, there may be an growing necessity for computerized, standardized, and scalable first-class warranty answers.
This research proposes an Artificial Intelligence and Machine Learning-based totally onion satisfactory classification gadget advanced the use of the TensorFlow deep gaining knowledge of framework. The gadget employs Convolutional Neural Networks (CNNs) to research pix of onions and classify them primarily based on observable attributes along with shape uniformity, surface texture, colour consistency, presence of mildew, sprouting, and symptoms of deterioration. The dataset used for training represents more than one pleasant conditions, together with fresh onions, partly broken onions, sprouted onions, and critically decayed onions. Preprocessing strategies such as photo normalization, contrast correction, and facts augmentation are implemented to make certain model robustness under real-world conditions. The CNN model is skilled using supervised studying, and its performance is evaluated using metrics including accuracy, precision, remember, and F1-score
The proposed system appreciably reduces the subjectivity and time required for onion satisfactory evaluation. By providing consistent and objective grading results, it enhances transparency in the agricultural deliver chain, supports truthful pricing mechanisms, and improves farmer earnings safety. Additionally, the system can be incorporated with cellular devices, automated sorting lines, and warehouse monitoring systems, permitting bendy deployment. The automation capability allows it to be used in both rural and industrial-scale environments. The research contributes to the developing course of virtual transformation in Indian agriculture and demonstrates how AI-driven inspection can lessen human dependency and economic inefficiencies. Overall, this work affords an reachable, price-effective, and realistic technique to improving onion pleasant grading through sensible photo-based analysis.},
        keywords = {},
        month = {January},
        }