Large-scale scene free bending pipeline detection and point cloud completion method based on four-dimensional space-time perception

Abstract

The invention discloses a large-scale scene free bending pipeline detection and point cloud completion method based on four-dimensional space-time perception, and the method comprises the steps: uniformly obtaining three-dimensional sampling points in a three-dimensional point cloud of a to-be-detected scene, building a spherical support domain, building a point pair type cylindrical feature set in a three-dimensional point cloud block, carrying out the second-class decision according to the feature clustering, and achieving the automatic detection of three-dimensional pipeline surface featurepoints; establishing a parameterized geometric model and a time sequence parameter estimation model of the pipeline in a four-dimensional time-space domain, solving a local pipeline section axis where pipeline feature points are located according to the geometric model to serve as a time domain initial state, and iterating towards the two ends of the axis at the initial moment in combination withthe time sequence parameter estimation model to carry out continuous pipeline section parameter estimation; obtaining a three-dimensional free bending pipeline surface point cloud fusing the three-dimensional geometrical information and the time domain related information and geometrical parameters thereof; reconstructing the pipeline surface point cloud according to the geometric model and iteratively optimizing to complete the pipeline point cloud in the original scene. The method has the characteristics of high calculation efficiency and point cloud noise resistance.

Description

technical field [0001] The invention relates to a feature detection and point cloud completion method of a free-bending pipeline facing a three-dimensional point cloud of a large scene, belonging to the field of industrial vision, in particular to the field of three-dimensional vision. Background technique [0002] In the field of industrial manufacturing and assembly, it is often faced with large-scale operation objects, such as aeroengines. The surface structure of the airframe is complex, and the types and quantities of parts are numerous and densely distributed. Each part plays an important role in the operation quality and safety of the airframe. The pipeline is an important part of the aero-engine. It is densely and staggeredly distributed in the aero-engine body. It mainly transports media such as oil and gas for the engine and the aircraft to ensure the stable operation of the engine. [0003] Due to the influence of temperature and pressure in the operating environ...

AI Technical Summary

Problems solved by technology

[0005] The technology of the present invention solves the problem: Aiming at the detection of three-dimensional free-bending pipelines on the body surface of industrial large-scale manufacturing parts, a large-scene free-bending pipeline detection and point cloud completion method with four-dimensional perception in time-space domain is provided, which can be adapted to be distributed in Pipelines with different thickness and bending degrees in complex large scenes can overcome the pipeline partition caused by pipeline connectors, realize the complete detection of multiple pipelines and complete the pipeline point cloud

Images