Cast pipe diameter identification method and device based on binocular vision

Abstract

The invention discloses a cast pipe diameter identification method and device based on binocular vision, and relates to the technical field of computer image processing and deep learning, and the method comprises the steps: obtaining a left view image and a right view image; performing image correction on the left and right view images; according to the corrected left view image or right view image, obtaining a sub Mask image containing the edge contour information of the inner diameter and the outer diameter of the cast pipe; fitting the inner diameter and the outer diameter of the Mask sub-image to obtain the center coordinate and the pixel diameter of the cast pipe; obtaining the distance from the cast pipe to the camera according to the corrected left and right view images; the real pipe diameter is identified according to the pixel diameter and the distance from the cast pipe to the camera, the identification method is not affected by temperature, water vapor and other environmental factors, the effects of inclination and shielding factors of the cast pipe are overcome, the accuracy and effectiveness of detection data are guaranteed, the processing speed is remarkably increased, and the method is suitable for popularization and application. And meanwhile, the portability and the stability of the method are greatly improved, so that the automation requirement of a cast pipe production line is met.

Description

technical field [0001] The invention relates to the technical fields of computer image processing and deep learning, in particular to a binocular vision-based cast pipe diameter identification method and device. Background technique [0002] With the rapid development of the Industrial Internet of Things, the automation and intelligence of the cast pipe production line has become an important development direction of the steel industry. Applying different equipment or leading to different production lines according to different diameters of cast pipes is the basis and key link to realize the automation of cast pipe production lines. The traditional pipe diameter recognition method uses laser recognition to divide the pipe diameter. The TOF algorithm used by laser recognition equipment relies on the condition of constant light speed to measure the transmission time of light to judge the distance. Therefore, there are temperature changes or water vapor in the use scene. The e...

AI Technical Summary

Problems solved by technology

However, in the real scene, since the ductile iron pipe rolls to the corresponding position in the non-driven track, the cast pipe will be tilted and the end surface position will fluctuate greatly, and the diameter of the cast pipe produced by the cast pipe production line will vary greatly, which will cause the traditional algorithm to fail. correctly identify and differentiate

In addition, the large difference in pipe diameter often causes the large pipe to block the small pipe, and the inclination of the end face further expands its shielding area. When the shielding area of ​​the end face is too large, the difficulty of identifying the pipe diameter will be greatly increased

Images