Beef and Pork Meat Classification using MobileNet V2 Algorithm via Android Smartphone Application

Abstract

Each type of animal meat exhibits distinct color and texture characteristics. For instance, beef typically presents a dark red hue with a chewy texture, while pork displays a paler red color with a smoother fiber. Previous research has utilized methods such as the gray level co-occurrence matrix (GLCM), hue saturation value (HSV), and color intensity for meat classification. In this study, we employed a MobileNet-V2 implemented in a Jupyter notebook, utilizing the MobileNetV2 model, to classify beef and pork meat. The dataset comprised 488 of each meat image after the augmentation process, partitioned into training (70%), testing (20%), and validation (10%) sets. Before partitioning, images were resized to 128×128 pixels. The model was trained using the training dataset with 100 epochs and the Adam optimizer, resulting in an accuracy of 96.93%.