Obstacle avoidance by sweeping robots is a commonplace topic. In the past two or three years, sweeping robot technology has developed rapidly, and sweeping robots have also become more and more intelligent from "mentally retarded". Next, we will introduce the mainstream obstacle

Obstacle avoidance of sweeping robots is a common topic. In the past two or three years, sweeping robot technology has developed rapidly, and sweeping robots have also become more and more intelligent from "mentally retarded". Next, we will introduce the mainstream obstacle avoidance methods of sweeping robots:

Sweeping robots The obstacle avoidance methods include: visual obstacle avoidance, laser radar obstacle avoidance, 3D structured light obstacle avoidance, 3D TOF obstacle avoidance

1

visual obstacle avoidance

binocular visual obstacle avoidance

is divided into monocular visual obstacle avoidance and binocular visual obstacle avoidance. Obstacle

• Monocular visual obstacle avoidance is to take photos through the camera mounted on the sweeping robot, and use machine learning - comparison - to identify obstacles. The disadvantage is the lack of depth information and two-dimensional pictures.

• The principle of binocular visual obstacle avoidance is similar to the human eye. The three-dimensional information of obstacles is constructed through two monocular parallax , which improves the recognition efficiency and provides better obstacle avoidance effect.

2

Lidar obstacle avoidance

Lidar obstacle avoidance

Lidar obstacle avoidance is to determine the distance and contour of the obstacle by emitting laser to the surface of the object and analyzing the intensity contrast of the reflected light, the length of time and other factors. The advantages are fast response speed, strong anti-interference ability, and adaptability to various light environments. It is much better than traditional infrared obstacle avoidance.

3

D structured light obstacle avoidance

3D structured light obstacle avoidance

D structured light obstacle avoidance has very high accuracy. It emits light with a certain structure through the sensor, captures the projected light spot, calculates the distance based on the deformation of the light spot, and collects the outline information of the obstacle. Use AI algorithms to identify and avoid obstacles.

4

D TOF obstacle avoidance

3D TOF obstacle avoidance

Its working principle is that the infrared light source emits high-frequency light pulses to the object, and then receives the light pulses reflected back from the object, and calculates the distance of the measured object from the camera by detecting the round-trip time of the light pulses.

Can the sweeping robot avoid obstacles 100%?

The answer is no. Current technology cannot achieve 100% obstacle avoidance.

summary: We only need to know that at the same price, binocular obstacle avoidance [like] is preferred, followed by the remaining monocular obstacle avoidance, lidar obstacle avoidance, 3D structured light obstacle avoidance, and 3D TOF obstacle avoidance.