Image Classification of Gamefowl Breeds Using MobileNet V2

Identifying specific gamefowl breeds poses a significant challenge due to their varied characteristics and the subjective nature of breed classification. This study aims to address these challenges by developing an image classification system using the MobileNet V2 algorithm to accurately classify gamefowl breeds based on their physical attributes. The model focuses on classifying four similar gamefowl breeds—Kelso, Lemon, Hatch, and Sweater—using full-body images captured by a USB camera, with real-time results displayed on an LCD screen. The study aims to assist breeders, enthusiasts, and poultry raisers in distinguishing between gamefowl breeds. The findings will provide valuable insights into the efficacy of MobileNetV2 for gamefowl breed classification, explicitly focusing on the chosen breeds.