Tap Wear Detection Based On Machine Vision

As a common tool for machining internal thread,the tap is more easily worn than other tools due to its bad tapping conditions and the structural characteristics of the tool itself,even the phenomenon of blade breakage and other failure.Usually,tapping is the final stage of the manufacturing process,so good tolerances and surface finish must be obtained.The wear degree of tap directly affects the thread quality and tool life.It is very important to measure the wear value of tap.The traditional measurement of tap wear is mostly measured by tool microscope,which is not only inefficient but also poor in stability.This paper takes tap as the research object,and through the method of machine vision,a set of tap wear detection system is designed and researched,which provides a new method for the tap wear detection.The main research contents and achievements of this paper include:(1)Based on the machine vision tap wear measurement model.The basic form and characteristics of the tap wear are analyzed,the blunt standard of the tap is introduced,and the wear of the tap is quantitatively analyzed.According to the characteristics of tap multi teeth,the automatic selection of tap teeth is realized by mode matching.The main back surface area is located and cut by the structure characteristics of the tap,and the wear value and the broken edge value are extracted from the segmented back cutter surface.(2)The construction of the tap wear detection system.According to the requirements of the system detection accuracy,the requirements analysis and selection of important hardware components which determine the image quality are carried out.In view of the difficult problem of image acquisition of the back of tap,special fixture and strip light source are used.The software part development tools use Lab VIEW and NI vision,and use MATLAB to design image processing algorithm,realize mixed programming,and finally design gui interface.(3)The design of the image processing algorithm for tap wear.Image processing includes preprocessing,cutting of main and rear cutter surface and measurement of wear value.In the aspect of filtering and noise reduction,the non local mean filtering algorithm based on noise estimation is adopted.In the aspect of single tooth segmentation,single tap teeth are located by NCC.The position of the main and rear cutting surface is deduced by using Hoff linear transformation to detect the auxiliary cutting edge of the bevel teeth.The wear area is divided by Otsu,and the broken edge area of tap is extracted by automatic multi threshold segmentation based on clustering,and the value of the broken edge is calculated.(4)The system of testing tap wear is tested by experiments.Through tapping experiment with tap,the error of the system measurement is less than 5%compared with the microscope measurement value.The effectiveness of the prototype system and the developed tool wear extraction algorithm is verified,and it can be used for automatic measurement of tap wear.Compared with the system and the super depth of field microscope and tool microscope,the system is more efficient and stable.