Simplified machine learning for image-based fruit quality assessment

Abstract

Fruit quality assessment is a crucial task in the fruit industry, traditionally done by human visual inspection. However, this process is subjective and time-consuming. This article proposes a simplified machine-learning approach for image-based fruit quality assessment. Our approach includes data collection, feature extraction using a pre-trained convolutional neural network, and classification using a support vector machine. We achieved an accuracy of 91%, precision of 92%, recall of 90%, and F1-score of 91%. Our approach can be applied to other fruits and integrated into automated fruit sorting systems, reducing the need for human inspection and improving the efficiency of fruit quality assessment.