Freshness of fruits and vegetables is an important factor in deciding their quality, and also important for attracting consumers with a high market price. Traditional methods for classifying fresh and rotten fruits and vegetables are time-consuming and dependent on a number of factors, such as size, shape, colour, etc. These methods also lead to inaccuracies and wrong classification. This paper presents a machine learning-based approach for automated classification of fresh and rotten fruits and vegetables. The proposed model uses feature extraction and k-Nearest Neighbors (k-NN) algorithm with Genetic Algorithm (GA), to improve accuracy and efficiency. We analyzed a dataset of 6,000 images (fresh and rotten) that includes images of fruits like apples, bananas, and oranges, as well as vegetables such as cucumbers, bitter gourds, and okra. The results demonstrate that the proposed model’s performance in classification accuracy, precision, and recall outperforms existing methods.