Research And Development Of The Visual Inspection System For The Hole Position And Size Of Iron Tower Components

The iron tower has large volume space truss structure made of angle steel and steel plate connecting members connected by bolts.It consists of numbers of steel plate linking components.There are many holes on the linking components.The accuracy of the holes on the connecting steel plates must meet the standard requirements to ensure the tower’s Installation quality.However,iron tower components are detected manually in current,which is difficult to adapt to the large scale of factory production.Therefore,it is great significance to develop a detection system which can detect the size and position error of hole machining position quickly,accurately and in batchAiming at the characteristics of large-scale processing and production of iron tower components and less inspection coverage samples,based on machine vision technology,the making-hole position and size detection system of iron tower components is researched.A set of visual inspection system for the hole-making position and size of iron tower components is developed to realize the hole processing of iron tower components automatic detection.By analyzing the actual environment and testing requirements for the iron tower components processing,the system architecture of testing plan is formulated which is suitable for the processing site.The system’s light source,lighting method,camera,lens selection,and other supporting components are analyzed and selected.The software development structure is determined.It is studied the structure of the software part of the system,and analyzed the role of each functional module of the system.The related algorithms of tower component feature image processing are researched.It is compared that detection effects of several different edge detection algorithms on the image.The sub-pixel edge detection algorithm based on Zernike moments are optimized to achieve image feature edge extraction.The least square method is used to fit the edge of the hole.The pixel clustering method is used to improve the Hough transform line detection algorithm and to solve incomplete recognition of outer contour line segments.Detection and comparison methods of the position and size of the feature hole are researched.The structure of the DXF file is analyzed;the size data extraction process is formulated.The feature data is converted from pixel coordinates to world coordinates,and further converted to machining coordinates.The comparison of hole size and position of iron tower components is finished based on DXF file.The experimental system is set up.The functions of each part of the system are verified.The error and repeatability experiments of single-type iron tower components were carried out.And the testing experiments of various types of iron tower components were completed.The average errors of hole position and hole diameter size were analyzed.The inspection efficiency was improved about 90% compared with manual inspection.The inspection system meets the requirements for accuracy,stability,applicability and detection efficiency.