Open-Box-Delivery Image Dataset

This dataset comprises a diverse collection of consumer product images categorized into four subsets \u2014 Clear, Blurr\/Dark\/Random, PrimaryPackage, and SecondaryPackage \u2014 designed for computer vision and image classification research. The images were collected under real-world conditions to represent varying lighting, clarity, and packaging contexts. The Clear subset (4,495 samples) includes high-quality images captured in optimal conditions; the Blurr\/Dark\/Random subset (1,476 samples) contains blurred, low-light, or non-uniform images representing challenging visual scenarios; the PrimaryPackage subset (1,701 samples) consists of images focusing on direct product packaging such as boxes or wrappers; and the SecondaryPackage subset (1,756 samples) includes images of outer or shipping packaging with labels and barcodes. This dataset supports the development and evaluation of robust deep learning models for object detection, packaging recognition, and visual quality assessment in retail and logistics applications.