Hybridization Methodology for Predicting ORE from Mines to End User using Machine Learning Technique

Iron ore is the most important iron ore in India and it is often lost on the way from the mine to its final destination. Manual sorting of many types of iron ore is difficult and time consuming. We deploy the most advanced RFID technology, along with weighbridges and GPS tracking systems, to solve this problem. Instead of using F-CNN, Iron ore grades were classified using a preliminary object recognition algorithm (YOLO V5). One crucial technical job in mining and excavation is locating mineral reserves. Conventional methods, however, are laborious and time-consuming. The concept makes use of machine learning capabilities to provide proactive, data-driven mine-to-end-user logistics, which minimizes environmental impact, lowers costs, and increases efficiency. Thus, datasets pertaining to mineral reserves were classified more accurately through the application of data augmentation and transfer learning. Comparative trials showed that using transfer learning may effectively counteract the improvement of categorization performance, up to and including 94%. This is a new concept for mining problems that can be solved by the above mentioned method.