Visual navigation uses the camera as the main sensor to sense the environment through visual information, save various environmental information as offline maps, and locate and navigate the robot based on the real-time image and offline map information matching when the robot is running.
HowADoes The Robot perceive the external environment?
The camera mounted on the robot can continuously capture images of the environment. We can extract identifying and unique information from it, including fixed objects such as walls and shelves, manual identification such as QR codes, and line segments and intersections. The geometric structure, this information constitutes the unique tags of specific objects in the external environment, just like various road signs and house numbers in the real world city, allowing mobile robots to determine where they are.
How Does A Robot Plan A Path?
After registering the offline map and the on-site construction drawing, given a specific target point in the map, we can calculate the optimal path on the map for a vehicle to the target point based on the current multi-vehicle operation conditions maintained by the system.
How Does The Robot Walk And Position?
The robot obtains real-time images from the on-board camera, extracts high-recognition information from them, and matches the information in the offline map to know its position on the map and in which direction it will move next.