Method for building semantic map for domestic robot and system

Abstract

The invention relates to a method for building a semantic map for a domestic robot. The method includes the following steps that: A, a two-dimension grid map of an indoor environment where the robot works is built; B, an RGD-D depth sensor is utilized to build a three-dimension map of the indoor environment where the robot moves; C, the types of furniture contacting a person are judged according to the behavior features of the person based on a wearable sensor; D, the built two-dimensional grid map and the three-dimensional map are fused, so that a three-dimensional indoor map can be built; and E, the types of the furniture in the indoor environment and the attributes of other markers are assigned into the three-dimensional indoor map, so that a three-dimensional indoor semantic map can be formed. According to the method of the invention, the types of the furniture in the indoor environment can be determined according to the movement of the person, and furniture information is added to the three-dimensional map, and therefore, the building of the three-dimensional indoor semantic map can be realized. The method has the advantages of structural simplicity, easiness in implementation, operational convenience and low cost.

Description

technical field [0001] The invention belongs to the field of intelligent semantic map construction, in particular to a method and system for building a semantic map by a household robot. Background technique [0002] Robots are used more and more widely, especially with the rise of mobile service robots, robots have begun to penetrate from traditional industrial fields to military, medical, home assistants and other fields. As the problem of social aging continues to intensify, the demand for mobile robots to enter the home to provide services for people is becoming more and more urgent. [0003] The environmental map created by the current SLAM (Simultaneous Localization and Mapping) technology has its limitations, that is, the built map is only valid for a period of time, and if the environment changes, it needs to be rebuilt. In addition, the existing map model only reflects the location information of the mobile robot itself, while the mobile robot wants to interact wit...

AI Technical Summary

Problems solved by technology

[0003] The environmental map created by the current SLAM (Simultaneous Localization and Mapping) technology has its limitations, that is, the built map is only valid for a period of time, and if the environment changes, it needs to be rebuilt

In addition, the existing map model only reflects the location information of the mobile robot itself, while the mobile robot wants to interact with people in the environment to realize functions such as anti-theft monitoring, cleaning, patient monitoring, control of home appliances, timekeeping reminders, and children's education. , then you need to use a semantic map, that is, you need to let the robot know its specific location, indicating whether the room is a "bedroom" or a "living room"

Images