Automated Walnut Classification and Weight Estimation Using Camera Images and ANN

Walnuts are a nutritious and delicious nut, as well as a nut that stands out with its economic value. Traditional methods are still widely used to harvest, dry, store, and finally market walnuts. However, these methods are often time‐consuming, costly, and require considerable expertise. Using image‐based deep learning classifiers to automatically classify walnut varieties can provide significant advantages in the agricultural sector and has the potential to revolutionize this field. The aim of this study is to automatically classify walnuts and predict their weights using camera images and artificial neural networks (ANNs). In the study, a dataset was created by collecting a total of 600 images from three different walnut varieties, namely Maraş 18, Chandler, and Bahri (Yayla). 70% of the samples were used as training data, 15% as validation data, and 15% as test data, and the network training was continued until the network performance dropped below 0.001%. The ANN model achieved 100% success in the classification of walnut varieties and 79% ( R 2 ) success in weighting. These results once again demonstrate the potential of artificial intelligence applications in agriculture. In future studies, it is recommended to test different artificial intelligence algorithms for the classification and weighting of different nuts.