An Edge-Based 3D Autonomous Exploration Method

Abstract

The invention discloses a three-dimensional independent exploration method based on the edge. According to the three-dimensional independent exploration method based on the edge, the edge is explored dynamically on the basis of exploration of the three-dimensional environment by a three-dimensional sensor and data fed back by the sensor each time by means of an unmanned aerial vehicle platform; traversal of all grids in a map is not needed when the edge is explored on the basis of map grid traversal, only the grids on the measurement contour of the sensor need to be traversed, in this way, calculated quantity is reduced, and calculation efficiency is improved greatly. The three-dimensional independent exploration method based on the edge is used for exploring several specific scenes and completing certain dangerous work without manual work and capable of greatly avoiding casualties.

Description

technical field [0001] The invention belongs to the fields of artificial intelligence and computer vision, and in particular relates to an edge-based three-dimensional autonomous exploration method. Background technique [0002] With the continuous development of research on intelligent robots, it has gradually become possible to use robots instead of humans to explore the environment. The key problem to be solved is the indoor autonomous exploration method for robots. [0003] The robot's indoor autonomous exploration refers to the robot's complete autonomous decision-making and pathfinding without human intervention, starting from the starting point, completing the exploration of the entire unknown environment, and recording the map information of the environment. The key issue is how the robot decides its next target point without human intervention. Usually such problems are solved using edge-based algorithms. [0004] The current indoor navigation of robots generally...

AI Technical Summary

Problems solved by technology

First of all, the environment where the robot lives is more messy and narrow, which increases the difficulty for the robot to avoid obstacles; secondly, in the indoor environment, there is generally no support from external navigation systems such as GPS, so it is impossible to directly obtain its own location information

[0008] Secondly, the traditional edge-based exploration strategy has the following deficiencies: First, since the original search strategy was proposed for two-dimensional ground robots, the sensor is a two-dimensional laser sensor, so the algorithm is only applicable to two-dimensional ground robots. Second, the search efficiency of the existing search strategies is low. Using edge detection and extraction algorithms in computer vision, it is necessary to master the information of the entire map and traverse all known and unknown area grids. , the time cost of the search is too high, and it is not suitable for rapidly changing environments and efficient exploration tasks; third, the existing algorithms cannot satisfy the dynamic edge search, and often need to fly to a target point, and collect environmental information after stopping , and then extract new edges, which will lead to the situation shown in Figure 2

Images