Simultaneous positioning and mapping SLAM is a technology used in robot motion. It means that a robot starts out in an unknown environment, locates itself through observed map features during motion, and then constructs a map based on its own position, thereby achieving the purpose of simultaneous positioning and map construction.

SLAM Flowchart

SLAM Core Issues

• Map building: how to integrate the information collected by sensors into a consistent model;
• Positioning: Estimate the robot's coordinates and posture on the map. SLAM will locate the robot on the map model while building a new map model or improving a known map.

SLAM Key Technologies

• Map representation
• Uncertain information processing method
• Data association
• Self-positioning
• Explore global path planning

SLAM Classification

According to the form and installation method of the sensor, it can be divided into two categories: lidar and vision:

LiDAR SLAM

• Advantages: It can measure the angle and distance of surrounding obstacles with high precision, high speed and small amount of calculation. It can be made into a real-time SLAM module. It is generally used to scan obstacles in a single plane, so it is suitable for robots with planar motion, such as unmanned vehicles and sweeping robots.
• Disadvantages: high manufacturing cost and relatively high price.

Visual SLAM

Visual SLAM is based on the growth of CPU and GPU processing speed and the improvement of hardware performance. According to the number and type of cameras, visual SLAM has three sub-directions, namely monocular, binocular and RGBD. In addition, there are special cameras such as fisheye and panoramic, but they are in the minority.