<![CDATA[Traditional methods for determining avocado ripeness rely primarily on subjective visual observation, which is highly prone to error. This study aims to develop an automated classification system for avocado ripeness using a sequential Artificial Neural Network (ANN) based on digital image data. The dataset consisted of 1,500 balanced avocado images distributed across three classes: Ripe (500 images), Rotten (500 images), and Unripe (500 images). A total of 1,200 images were used for training and 300 images for validation. Images were preprocessed through resizing to 128 × 128 pixels and pixel intensity normalization. The proposed ANN architecture consisted of a Flatten layer, two hidden Dense layers with ReLU activation, and an output layer with Softmax activation. Experimental results showed that the model achieved an overall accuracy of 89.00%, with macro-average Precision, Recall, and F1-Score values of 0.92, 0.89, and 0.89, respectively. The best classification performance was achieved for the Rotten class, with a Precision of 1.00 and a Recall of 0.96. Classification errors mainly occurred between the Unripe and Ripe classes, where visual similarities during the ripening transition stage led to cross-class predictions. Overall, the proposed ANN model demonstrated reliable performance for avocado ripeness classification using digital image data and showed its potential as a simple image-based decision-support tool for post-harvest quality assessment.]]>
