CN-117047313-B - Autonomous feeding method, device and medium for special pipe

CN117047313BCN 117047313 BCN117047313 BCN 117047313BCN-117047313-B

Abstract

The embodiment of the specification discloses an autonomous feeding method, equipment and medium for a special pipe, wherein the method comprises the steps of obtaining a first cross-section outline drawing of the special pipe at a computer end, calculating a first rotation vector of the first cross-section outline drawing, obtaining a second cross-section outline drawing of the special pipe to be processed based on an industrial camera, calculating a second rotation vector of the second cross-section outline drawing, calculating similarity scores of the first cross-section outline drawing and the second cross-section outline drawing, adjusting parameters of the industrial camera when the similarity scores are lower than a preset threshold value, repeatedly executing the steps of obtaining the second cross-section outline drawing until the similarity scores are higher than the preset threshold value, and obtaining a rotation angle of the special pipe to be rotated based on the first rotation vector and the second rotation vector when the similarity scores are higher than the preset threshold value, so that a mechanical chuck resets the special pipe to be processed based on the rotation angle. Autonomous feeding of the pipe laser cutting equipment is realized, and production efficiency is improved.

Inventors

Zhang Qinmi
JIAO YUPENG
LI QIANG
SUN MINGJIE
CAO HAOWEI

Assignees

济南邦德激光股份有限公司

Dates

Publication Date: 20260508
Application Date: 20230921

Claims (6)

1. An autonomous feeding method for a special pipe is characterized by comprising the following steps: acquiring a first cross-section outline drawing of a computer-end special-shaped tube, and calculating a first rotation vector of the first cross-section outline drawing; acquiring a second cross-sectional profile of the special pipe to be processed based on an industrial camera, and calculating a second rotation vector of the second cross-sectional profile; Calculating a similarity score for the first cross-sectional profile and the second cross-sectional profile; when the similarity score is lower than a preset threshold, adjusting parameters of the industrial camera, and repeatedly executing the steps of obtaining the second cross-sectional profile map to calculate the similarity score until the similarity score is higher than the preset threshold; When the similarity score is higher than the preset threshold value, a rotation angle of the special pipe to be rotated is obtained based on the first rotation vector and the second rotation vector, so that the mechanical chuck resets the special pipe to be processed based on the rotation angle; The method for acquiring the first cross section profile of the computer-end special-shaped tube specifically comprises the following steps: acquiring a cross section engineering drawing of a special pipe in a computer end, and extracting outline size information in the cross section engineering drawing; Converting the cross section engineering drawing into an initial cross section picture in a picture storage format; calculating a picture pixel scale in the initial cross-section picture according to the camera pixel scale of the industrial camera and the outline size information; scaling the initial cross-section picture to obtain a target cross-section picture with the camera pixel scale consistent with the picture pixel scale; the calculating the first rotation vector of the first cross-sectional profile specifically includes: if the figure outline in the target cross-section picture shows rotational symmetry, calculating the rotation angle between adjacent symmetry axes in the rotational symmetry figure as the angle of the rotation vector; If the figure outline in the target cross section picture does not show rotational symmetry, calculating a vector of a rotation center in the target cross section picture, which points to the outline centroid, as the first rotation vector; The obtaining of the second cross-section profile map of the special pipe to be processed based on the industrial camera specifically comprises the following steps: extracting a preset area from an initial image acquired by an industrial camera to obtain a cross section area of the special pipe between clamping jaws; adopting a bilateral filtering function to perform noise reduction treatment on the cross section area of the special-shaped tube to obtain a noise reduction image; performing exposure treatment on the noise reduction image, extracting contour edges under different exposure degrees, and fusing to obtain an image contour map; and removing pixel information except the contour information in the image contour map to obtain the second cross-section contour map.
2. The method for autonomous feeding of a special pipe according to claim 1, wherein the extracting the predetermined area from the initial image obtained by the industrial camera to obtain the cross-sectional area of the special pipe between the clamping jaws comprises: identifying a jaw profile in the initial image; Determining the maximum circumscribed rectangle of all the outlines in the clamping jaw outline; and taking the pixel area in the maximum circumscribing rectangle as the cross section area of the special-shaped tube.
3. The method for autonomous feeding of a profile tube according to claim 1, wherein the calculating the similarity score of the first cross-sectional profile and the second cross-sectional profile specifically comprises: cutting the target cross section picture to obtain a template image; Calculating an angle difference value between the first rotation vector and the second rotation vector, and rotating the template image according to the angle difference value; training the rotated template image, performing template matching on the rotated template image and the second cross-sectional profile obtained by the industrial camera, and calculating to obtain a similarity score.
4. A method for autonomous feeding of a profile tube according to claim 3, wherein said training of said rotated template image comprises: Setting a first angle threshold and a second angle threshold as angle training parameters according to an initial angle corresponding to the initial state of the template image, and setting an angle training step threshold; extracting pyramid images corresponding to the template images and extracting image feature points of each layer of pyramid images based on the size of the template images; according to the image feature points, the angle training parameters and the angle training step length, carrying out gradient feature training on the template image to obtain matching data; and saving the trained matching data.
5. Autonomous feeding equipment for special pipes is characterized in that the equipment comprises: At least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores instructions executable by the at least one processor, the instructions are executable by the at least one processor to enable the at least one processor to: -performing a method for autonomous loading of tubes according to any of claims 1-4.
6. A non-volatile computer storage medium storing computer executable instructions which, when executed, implement a method of autonomous loading of a tube according to any one of claims 1 to 4.

Description

Autonomous feeding method, device and medium for special pipe Technical Field The specification relates to the technical field of laser cutting, in particular to an autonomous feeding method, device and medium for a special pipe. Background Along with the continuous promotion of research and development level, the intelligent degree of laser cutting equipment is higher and higher. To date, laser cutting equipment has replaced the manual work to realize the autonomous material loading of specific mechanical tubes. However, the laser cutting device has strict requirements on the initial pose of the special tube when the special tube is machined, and because the existing laser cutting device cannot accurately control the feeding pose of the special tube, the machining of the special tube in the laser cutting device needs to be manually assisted in feeding. The manual auxiliary feeding has the following defects: 1. The time taken to manually adjust the profile tube is relatively long, which directly results in a low production efficiency. 2. When the special pipe is manually adjusted, an operator is required to manually contact the running laser cutting equipment, and the safety of the operator cannot be completely ensured. 3. For the special-shaped tube which is not easy to place by the chuck, the position and the posture of the special-shaped tube cannot be reset accurately manually. 4. Inaccuracy in machining the special-shaped tube can cause waste of special-shaped tube materials. 5. The laser head and the position of the processing plane have errors, so that the phenomenon of scraping and touching the laser head can occur, and the service life of the laser head can be directly reduced. Due to the problems, the existing laser cutting equipment cannot accurately and autonomously cut the special pipe with higher complexity. The autonomous feeding of the special pipes in the laser cutting equipment is still blank in the domestic market. According to the application, the pose recognition and intelligent decision of computer vision are integrated into the laser cutting equipment, so that the autonomous feeding of the pipe laser cutting equipment is completely realized, the problem of cutting precision reduction caused by manual errors is avoided, the resetting efficiency is improved, the labor cost is reduced, and the production efficiency of the laser cutting machine in the pipe cutting process is improved. Disclosure of Invention One or more embodiments of the present disclosure provide a method, an apparatus, and a medium for autonomous feeding of a profile tube, which are used for solving the technical problem that an existing laser cutting apparatus cannot accurately and autonomously cut a profile tube with higher complexity. One or more embodiments of the present disclosure adopt the following technical solutions: one or more embodiments of the present disclosure provide a method for autonomous feeding of a profile tube, the method comprising: acquiring a first cross-section outline drawing of a computer-end special-shaped tube, and calculating a first rotation vector of the first cross-section outline drawing; acquiring a second cross-sectional profile of the special pipe to be processed based on the industrial camera, and calculating a second rotation vector of the second cross-sectional profile; Calculating a similarity score for the first cross-sectional profile and the second cross-sectional profile; when the similarity score is lower than a preset threshold, adjusting parameters of the industrial camera, and repeatedly executing the steps of obtaining the second cross-sectional profile map to calculate the similarity score until the similarity score is higher than the preset threshold; And when the similarity score is higher than the preset threshold value, obtaining a rotation angle of the special pipe to be rotated based on the first rotation vector and the second rotation vector, so that the mechanical chuck resets the special pipe to be processed based on the rotation angle. Further, the acquiring the first cross-section profile of the computer-end special-shaped tube specifically includes: acquiring a cross section engineering drawing of a special pipe in a computer end, and extracting outline size information in the cross section engineering drawing; Converting the cross section engineering drawing into an initial cross section picture in a picture storage format; calculating a picture pixel scale in the initial cross-section picture according to the camera pixel scale of the industrial camera and the outline size information; And scaling the initial cross-section picture to obtain a target cross-section picture with the camera pixel scale consistent with the picture pixel scale. Further, the calculating the first rotation vector of the first cross-sectional profile specifically includes: if the figure outline in the target cross-section picture shows rotational symmetry, calculating the rotati