Visual inspection method and system for three-dimensional bent pipe

Abstract

The invention discloses a visual detection method and system for a three-dimensional bent pipe, and the method comprises the steps: collecting an image of a to-be-detected bent pipe through a three-dimensional reconstruction module, obtaining single bent pipe point cloud data, and carrying out the visual positioning of the current spatial position and posture of the three-dimensional reconstruction module through a visual positioning module; fusing the single bent pipe point cloud data according to a visual positioning result to form complete bent pipe point cloud data; recognizing the cylindrical section and the arc section of the bent pipe according to the complete bent pipe point cloud data, and determining parameters of the cylindrical section and the arc section of the bent pipe. Therefore, the three-dimensional bent pipe workpiece can be rapidly and accurately detected, and the method and the system have the advantages of being convenient to operate, high in flexibility degree and accurate in measurement.

Description

technical field [0001] The invention relates to bent pipe detection, in particular to a visual detection method and system for three-dimensional bent pipes. Background technique [0002] The main methods of traditional industrial elbow inspection include: checking tool inspection (profiling method), visual inspection based on two-dimensional images, forming measurement, fork measurement and other methods. Among these existing detection methods, the operation of the inspection tool detection method is time-consuming, full inspection cannot be achieved, the mold volume is large and it is not convenient to store, the results are poorly traceable, and it is difficult to quantify; The requirements are high, and some testing equipment are large in size, which cannot adapt to the occasion where the elbow is not easy to disassemble; the forming measurement method is not efficient, and the fork measurement method is only suitable for low-precision measurement occasions; the joint arm...

AI Technical Summary

Problems solved by technology

Among these existing detection methods, the operation of the inspection tool detection method is time-consuming, full inspection cannot be achieved, the mold volume is large and it is not convenient to store, the results are poorly traceable, and it is difficult to quantify; The requirements are high, and some testing equipment are large in size, which cannot adapt to the occasion where the elbow is not easy to disassemble; the forming measurement method is not efficient, and the fork measurement method is only suitable for low-precision measurement occasions; the joint arm, For high-precision machinery such as robotic arms, the measurement range is limited for the measurement of large-sized pipe bends. Secondly, mechanical equipment such as articulated arms are not portable enough, so there are certain requirements for the measurement scene.

Images