Coffee Beverage Product Occlusion Image Dataset

The retail e-commerce industry faces challenges with product visibility due to occlusions from other items, which can impair customer decision-making. Current solutions often rely on traditional product images without considering occlusion, leading to inaccuracies in product identification and customer dissatisfaction. This dataset aims to address the technical issue of occlusion in product images by providing a comprehensive collection of occluded coffee beverage product images, allowing for better training of machine learning models for recognition tasks. Data was collected using high-resolution cameras in controlled environments, ensuring clarity even with occlusions. Quality control was implemented through multiple rounds of annotation, consistency checks, and expert reviews. The dataset is organized in JPG format, with images stored in a structured directory based on categories and occlusion levels. Each image is linked to metadata containing the occlusion degree and product labels.