Warehouse automation is the automating practice of inventory control in-and-out of warehouses for consumers with zeroed human support. An enterprise can exclude labor-intensive costs which implicate Intelligent Storage System (ISS), Clustering and Racking Systems (CRS), and data analytics. The enterprise can run CRS manually, but it is extremely complex and can lead to an error-prone. This paper thus introduces the computer vision, upon Artificial Intelligence (AI) based training models with digital images and deploys to on-the-site devices that detect and interpret visual perception of products from cameras. Machine learning with visual AI recorded images of containers/products is used to equip the system with cognitive skills in the warehouse. In brief, it recognizes what to place into the warehouse, what checks out, where it is its cluster, and where it may have been relocated. This helps prevent problems arise in the racking system, where it could over-sit beyond expiry and become worthless. Both supervised and unsupervised training approaches for CRS with digital images are simulated. Finally, as a result, the computer vision speedily tracks warehouse products regardless of RFID tags, which subsides the limitations of RFID tags like vanished tags, imperfectly assigned tags, broken or non-functioning tags and the cost of re-issue tags.