A Fast Warehouse Inventory MAV system Equipped with Visual Guidance and Navigation Algorithm

This paper is an attempt to integrate computer vision techniques and MAV guidance to design and optimize an automated mission performed by a light MAV such that automating the mission becomes reasonably more efficient than performing it manually. A system is provided for warehouse management using an MAV equipped with a front camera. Computer vision algorithms makes it possible for the MAV to locate packages, verify the presence or absence of a specified package and list the entire warehouse inventory in a short time. An innovative method is provided to detect shelves and their packages by the camera image, which enables the system to instantly plan the shortest path for the MAV while performing a shelf inventory listing. Then, following the planned path completes the mission faster than conventional guidance methods. The guidance algorithm is designed such that the efficiency of automatic operations compared to human operations is significantly increased. The system is first simulated and then implemented and the test output data is provided. The tests indicate the success of the system in securing automated operations, while decreasing mission time.