Bag Damage Detection Dataset

Description

The "Bag Damage Detection" dataset focuses on damages in bag, providing a specialized collection of 700 images captured using an iPhone 13 12MP rear camera. This dataset was gathered between October and December 2023, from the India Innovation Centre Vanderlande in Pune, as well as from platforms like Roboflow, YouTube, iStock, and Shutterstock. The images capture various scales and angles, simulating real-world conditions encountered in baggage handling systems. Each image is annotated with segmented bounding boxes marking the precise locations of damages such as tears, rips, and punctures, labeled with the classes "damage" and "bag." This detailed labeling will support the development of instance segmentation models using the latest algorithm. The dataset undergoes preprocessing steps including orientation, resizing to 640x640 pixels, and augmentation techniques like flips, rotations, and noise addition to enhance the model's robustness and generalization capabilities. This dataset is critical for applications in automated baggage handling, quality control, and security screening, aiding in the development of systems that can accurately assess the condition of luggage, thereby improving operational efficiency and reducing manual inspection efforts.

Categories

Funding

Licence