Lidar YOLO 3D MAP 3D MAP self-driving drone combines Lidar sensor with YOLO algorithm

摘要

The present invention relates to a 3D MAP autonomous driving vehicle that combines a lidar sensor and YOLO algorithm, which can have higher accuracy than an autonomous driving method that relies on existing LIdar sensors or cameras. By applying the object recognition algorithm to drones to show the many possibilities of 3D mapping and to make it applicable to many businesses, it is intended to improve the inefficient use of the existing data processing process and solve the problem of increasing the accuracy that is lacking with existing sensors. In other words, in the present invention, the Jetson Nano is used for high computational power of the YOLO algorithm so that the information received from the lidar sensor can be combined with the information recognized by the YOLO object recognition algorithm in a wireless vehicle using an advanced sensor to exhibit higher accuracy. The YOLO algorithm, which is composed by combining the MCU and LIDAR sensor with the flight controller PIXHAWK, after supporting computational power by combining a high-performance external MCU, and the YOLO object recognition algorithm. Applying this high Jetson Nano as an MCU After combining it with the drone frame to process the signals received from the external MCU (Jetson Nano), MCU and the drone controller, which consists of integrated control of different software and hardware with ROS, the MCU and drone radio waves Flight Controller (PixHawk) with a wired connection to the receiver, a propeller combined with a motor to make the drone fly, a motor combined with a propeller and drone frame to move the drone, and a drone body, flying and landing to assemble all components A drone frame that combines gears, a GPS attached to a drone frame to collect location information and wired to FC, a transmitter attached to a frame to receive external radio waves and wired to an MCU, and an FC to collect external information. It consists of a lidar sensor configured to send a signal to the MCU after connection, and a radio wave receiver that is attached to the drone frame to receive a signal from the drone controller and connected to the FC by wire. Therefore, the present invention can have higher accuracy than the autonomous driving method that relies on the existing LIdar sensor or camera, and many possibilities of 3D mapping by applying the Lidar sensor and YOLO object recognition algorithm, which are currently only used in 2D mapping of autonomous vehicles, to the drone. It has the effect of improving the inefficient use of the existing data processing process and solving the problem of increasing the accuracy that is lacking with the existing sensor by making it applicable to many businesses.