A two-dimensional laser real-time detection method of workpiece surface contour

Abstract

The present invention provides a two-dimensional laser real-time detection method for the surface contour of a workpiece. For the contour point data of the environment or the target object acquired by the 2D laser sensor, the contour point data is first over-segmented using the adaptive threshold IEPF algorithm; and then Construct an undirected graph, the over-segmented point set data is used as an undirected graph node, the probability of fusion of the over-segmented point set is used as an edge of the undirected graph, and the fusion probability value is calculated; then the energy function for segmenting the undirected graph is constructed and the energy function is proposed Minimize the solution to obtain the fusion result of the over-segmented point set, that is, the line segment fitting result of the point data; finally, use the prior knowledge and the shape template to calculate the line segment fitting result to obtain the shape detection result; this method overcomes the traditional algorithm. Threshold sensitivity, sensitivity to data noise, and weak robustness can be used for the detection of specific-shaped workpieces of industrial robot arms, scene understanding of autonomous moving robots, and environmental cognition of unmanned vehicles.

Description

technical field [0001] The invention relates to the technical fields of 2D laser sensors, laser radar sensors, signal processing, and undirected graph clustering, in particular to a two-dimensional laser real-time detection method for workpiece surface contours. Background technique [0002] For industrial robotic arms, autonomous mobile robots, unmanned vehicles and many other fields, environmental perception is a must-have basic capability. Different sensors are used for environment perception by mimicking the human perception mechanism. Among them, the laser sensor is widely used in various systems because of its high measurement accuracy and good real-time performance. Industrial robotic arms mostly use 2D laser sensors, while autonomous mobile robots and unmanned vehicles mostly use multi-line lidar. The key to environmental perception is to use laser sensors to scan the environment or targets, obtain the contour distance data of the environment or targets, and extrac...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above-mentioned problems in the prior art, the object of the present invention is to provide a two-dimensional laser real-time detection method for the surface contour of the workpiece, which solves the problem that the current common methods cannot extract straight line features from laser data stably and robustly. problem, and use this method in the shape detection of laser sensor contour data to provide a basis for subsequent data processing, decision-making and control

Images