Inventory Management Flow Tracking

• Implementing object detection and tracking algorithms based on Faster R-CNN to monitor the inventory I/O flows in warehouses by surveillance footage, which automates cargo management process to reduce artificial errors and human costs.

• Optimizing a real-time tracking algorithm by enhancing focus attention mechanism with background subtraction in OpenCV.

• Designing a multi-camera cross-view data fusing algorithm to increase the coverage of visible areas and detection recall.

• Constructing training datasets containing 30,000 images of objects in warehouses from raw footages provided by ThaiBev