Study On The Detection System Of Inner And Outer Wall Of Small-diameter Workpiece Based On Machine Vision

In recent years, micro-nanofabrication technology continues developingin China. Micro machines with small size have been highly developed andwidely used in the field of precision instruments, machinery manufacturing,aerospace and others. Therefore, the detection of size and quality of suchmicro mechanical work pieces is highly demanded. Especially, the detectionof inner wall of the small-diameter work piece is always a hard problem.Industrial endoscope method and eye-inspection method are the universalsolutions to this problem, but it is difficult to ensure the measurementrepeatability and it cannot save images. These both methods waste time andenergy to some extent. If there are defects in the inner wall surface of thework piece, they will not only affect the appearance of the work piece, butalso affect the function of it greatly, like its stability and security.Firstly, the existing detection methods of the detection of inner wall ofthe small-diameter work piece are discussed and analyzed. Then it analyzestheir main strengths and weaknesses, and it is to put forward a new kind ofmachine vision detection method based on the mirror. The system aims atone kind of small-diameter (2~5mm) work piece, which is with a shallowthrough hole. It gives research and analysis of three detection programs ofthe inner and outer walls of three small parts. Through feasibility analysisand experimental basis, the final measurement program is determined, themechanical structure of the system is designed and built, and an entirehardware system and a preliminary design of software system are completed.This kind of detection system and program proposed in this paper isbased on an inner mirror, getting the image by the use of the principle oflight reflection. It uses digital image processing techniques to get the complete image of the whole inner wall. This system can effectively obtainimages of inner and outer walls of the small-diameter. The operation ofprogram is simple and it is easy to save images. The obtained images’resolution and repetition rate are both high.