Akademik Veri Yönetim Sistemi
International Journal of Agriculture, Environment and Food Sciences, cilt.9, ss.82-91, 2025 (TRDizin)
Advancements in unmanned aerial vehicle (UAV) technologies have facilitated a novelapproach to dairy cattle breed morphological identification. The objective of this studywas to employ UAV images, analyzed through deep convolutional neural networks(DCNN), to classify dairy cow breeds. The dataset comprises of 2004 RGB UAV imagesof dairy cows, including Holstein, Simmental, and Brown-Swiss breeds, obtained fromthe cattle breeding facility at Van Yüzüncü Yıl University. The images were preprocessedand segmented to contain a single cow each, and subsequently categorized as training(70%), validation (20%), and testing (10%) datasets. To determine the most effectivearchitecture for breed classification, we compared a custom DCNN (C-DCNN) model towell-established pre-trained models including Xception, VGG19, and ResNet50. The C-DCNN demonstrated remarkable performance, achieving precision, recall, accuracy, andF1 scores of 0.98. Among the pre-trained models, Xception demonstrated superior results,with perfect accuracy and an F1 score of 1.00. Conversely, the VGG19 model exhibiteda higher level of accuracy; nevertheless, it exhibited lower precision, recall, and F1 scoreswhen evaluated on the test set, compared to the C-DCNN and Xception models. Incontrast, ResNet50 displayed the lowest level of performance, with an accuracy of 0.74and the highest levels of loss. This study demonstrates the potential of integrating DCNNmodels with UAV technology in precision livestock farming, offering a robust andefficient system for cattle breed classification.