YOLOv8 Powered Solutions for Box Identification in Warehouses

Abstract

In modern warehouse management, the ability to effectively identify and track boxes is critical for optimizing operations and reducing costs. This research investigates the application of YOLOv8 deep learning model for real-time box identification in warehouse environments. Three different approaches were evaluated: using a pre-trained YOLOv8 model, training the model with a dataset obtained from the Internet, and training the model with a custom dataset designed for this application. For the second and third approaches, the model was trained using Google Colab, and image annotation was performed using Roboflow. Each approach is thoroughly tested to assess the accuracy and robustness of the model under various conditions. The results demonstrate the strengths and limitations of YOLOv8 in different scenarios, providing valuable insights into its practical implementation for warehouse automation. This study highlights the potential of YOLOv8 as a useful tool for improving warehouse efficiency.