Research On Embedded Small Modular Gear Defect Visual Inspection System

In recent years,with the rapid development of computer technology and digital image processing technology and the continuous improvement of related algorithm theories,profound change has taken place in the detection technology of surface quality of products.The research level in the field of detecting the surface quality of small modulus gear has become more deeply,the detection methods has become more intelligence.Small module gears play an important role and are widely used in the field of mechanical transmission.Once the defective products flow into the market,it will cause some adverse effects to users and gear manufacturers.One of the means to improve the gear quality is making use of various detection techniques to detect the defective products that appear in the manufacturing process and remove them in time.The traditional method of product surface inspection depends heavily on the artificial vision,however,have high detection costs and poor stability.Some of detection methods,such as magnetic particle testing and eddy current testing,cannot be used to detect the plastic material.The Machine Vision Inspecting technology,which is an advanced method for testing the surface quality of products,has the incomparable advantages over other detection methods in the aspects of efficiency,intelligence,accuracy and reliability.It has become the popular research method at present.The vision technology based on PC-board system has been already studied and applied in gear surface quality detection.However,there are still some problems,such as high testing cost,large equipment volume and inconvenient maintenance.With the advancement of microelectronics and very large scale integrated(VLSI)manufacturing technology,embedded technology has also been increasingly applied in social life.It is an important trend to develop an embedded vision system with low cost,high flexibility and low power consumption in product surface quality detection.This paper has completed the defect detection of the surface quality of small modulus gear products through the embedded technology and visual inspection technology,and improves the quality of the gears out of the factory.The main work finished in the thesis is showing as follows:(1)The article analyzes the status quo and trend of development of small module gear detection technology,discusses some common detection methods of the current product surface quality,expounds the principle of machine vision detection and analyzes the basic components of embedded vision system.By the comparison with PC-Based vision system,some reasons are given that why choose the embedded visual inspection system.(2)Some key technologies needed in the process of system design are discussed,and the platform features of the embedded system are introduced.The advantages and disadvantages of the hardware platform of embedded processor based on FPGA,ARM and DSP respectively in the development process are compared.What's more,different embedded systems and visual toolkits are also compared and some core detection algorithms used in this paper are discussed.(3)According to the characteristics of small modulus gear,related researches on image processing methods such as enhancement,segmentation and feature extraction are carried out.The simulation results indicate that the proposed method,which mainly makes use of the speeded up robust features(SURF)and image difference,has proved a good detection effect on the surface quality detection of small modulus gear.(4)With simple operation,friendly interface and high stability,a graphical user interface(GUI)program has been designed.The relevant core algorithms involved in this article are implemented in the C++ programming language.Through the Qt image application programming interface,the acquisition and analysis of the related images is finished.Then the embedded transplantation of the whole detection system is completed.With the related research works in this thesis,the embedded small module gear defect visual inspection system designed in this paper owns the characteristics of low cost,strong real-time performance and high reliability,having good practical value and market prospect.