Short-range obstacle avoidance method and short-range obstacle avoidance system based on three-dimensional depth camera

Abstract

The invention provides a short-distance obstacle avoiding method based on a three-dimensional depth camera, which comprises the steps of acquiring a depth image ahead when a moving device moves, wherein the depth image comprises ground data and aerial data; when existence of an obstacle on the ground in the depth image in front of the moving device is detected, calculating an angle and a distance between the current position and the obstacle, and changing a moving path of the moving device according to the calculated angle and the calculated distance; judging a ground portion without data is the obstacle when it is detected that part or all of the ground data in front of the moving device cannot be acquired, calculating an angle between the current position and the obstacle, and changing the moving direction of the moving device according to the calculated angle. The invention further relates to a short-distance obstacle avoiding system based on the three-dimensional depth camera, thereby being capable of realizing short-distance obstacle avoidance safely.

Description

technical field [0001] The invention relates to the field of electronic technology, in particular to a short-range obstacle avoidance method and a short-range obstacle avoidance system based on a three-dimensional depth camera. Background technique [0002] Mobile robots perceive the environment through sensors, but the current sensor's ability to perceive the environment is not perfect, the data given by the sensor is inaccurate, and there are blind spots in the sensing range. Among the existing sensors, the 3D depth camera is a better choice, which has high accuracy in acquiring the environment, but there is a problem of sensing blind spots. The current imaging principles of 3D depth cameras can be roughly divided into two types: one is to obtain depth through structured light, and the infrared laser emits relatively random but fixed spot patterns. The positions captured by the camera are also different, and then calculate the displacement of the spots of the captured pat...

AI Technical Summary

Problems solved by technology

[0003] The data acquired by the 3D depth camera is accurate, but due to the accuracy of the equipment itself, there is no method of using the 3D depth camera for 3D obstacle avoidance in the existing obstacle avoidance methods.

Because the 3D depth camera has short-distance perception blind spots, it cannot measure small displacements or small time changes, making it impossible to obtain data on objects that are too close. When obstacles are in the blind spot, it will be impossible to safely avoid obstacles, such as the kinect generation of 3D depth cameras. The range of short-distance perception of the environment is 0.4m to 3.5m, and the range of 0m to 0.4m is the working blind area of ​​kinect. When an object approaches the robot’s working blind area or suddenly appears in the robot’s working blind area, kinect will not be able to find obstacles, but cannot Realize safe obstacle avoidance, which is a big flaw in how mobile robots can safely and autonomously avoid obstacles

[0004] If it can overcome the defect that the 3D depth camera cannot achieve close-range safe obstacle avoidance, it will be of great significance to improve the 3D obstacle avoidance

Images