Fair Price Prediction for Farmers: Leveraging Freshness of Perishable Goods Through Machine Learning and Sustainable Practices

Fresh produce enhances sales and promotes better health, whereas stale stock leads to financial losses and significant food wastage. Additionally, inefficiencies in the perishable food supply chain contribute to sustainability challenges. This study proposes a system that integrates freshness classification and fair pricing mechanisms to support farmers and mitigate waste. The approach employs the VGG16 model to classify produce as fresh or rotten, achieving 96.88% accuracy. Further freshness analysis is conducted using image processing techniques such as edge analysis with the Laplacian of Gaussian (LoG), texture analysis with Local Binary Patterns (LBP), and color analysis using the HSV model, yielding accuracies of 82%, 86%, and 87%, respectively. A computed freshness index enables shelf-life estimation, helping optimize inventory management and reduce waste. Additionally, a fair pricing model based on the freshness index ensures that farmers receive equitable compensation for their produce. The SARIMAX model, used for price prediction, achieves a Mean Absolute Percentage Error (MAPE) of 28.18%. By linking freshness assessment with pricing, this framework incentivizes better agricultural practices, enhances efficiency, promotes sustainability, and minimizes environmental impact. The proposed system contributes to a resilient and eco-friendly perishable food supply chain, benefiting both farmers and consumers.