This smart retail cabinet technology integrates deep learning and computer vision, achieving accurate product recognition in complex environments through super-resolution, anti-glare, and data augmentation methods. It intelligently locates product row and column positions on shelves based on clustering technology and uses fine-grained recognition to distinguish subtle differences between highly similar products. The system supports the closed-loop management of standardized product placement, automatic inventory counting and smart sales.

Technology provider:Nanjing University of Posts and Telecommunications