Study On Pitch Measurement System For Transmitting Signal Plate Based On Machine Vision

With the development of machinery industry, the shape of the work piece is moreand more complex; meanwhile, the requirements of accuracy and detection speed arehigher and higher. Traditional manual measurement methods have been unable to meetthe actual needs of the modern industry. Transmitting Signal Plate is one of the mostimportant components in automobile engine. It has great demand and high productionrequirements. Transmitting Signal Plate itself has a typical gear-like structure. Workersmust spend a lot of time on measuring those teeth course of its poor efficiency and lowstability. This kind of measurement method is not suitable for the automatic productionline. Therefore, the improvement of the measurement method is a hot issue.Machine vision technology as the rising technology for decades combines computerscience with high-speed computational capability and electronic imaging system with2-dimentional imaging capability. The system simulates parts of human visualperformance and develops some new capabilities that human being don’t possess.Measurement techniques based on machine vision is one of them. It was developedbased on measurement technique and computer vision as a new interdisciplinary. Thisissue introduced measurement based on machine vision to the Pitch MeasurementSystem of Transmitting Signal Plate and studied hardware structure, software andalgorithms systematically.The main content of the study includes the following aspects:(1) Machine vision measurement development at home and abroad, explores thebasic principles and main methods of the Pitch Measurement System of TransmittingSignal Plate；(2) Carried out a systematic study of all aspects of the algorithms in the processingof visual measurement system software and introduced the algorithm used in this thesis;The dimensionality reduction method with coordinate transformation was proposed toextract points of interest from the two-dimensional image to a Cartesian coordinatesystem, reducing the amount of computing; Meanwhile, variation range least squaresfitting method used in sub-pixel edge computing was also proposed in order to make therequirements of the number of variables with greater flexibility when calculating thesub-pixel edges, accelerate the speed of production while strengthening the qualitymonitoring.(3) Analyze different parts of hardware system, select and set up an appropriatehardware platform of transmitting signal plate image capture. Using Visual Studio2010as Windows application development platform to build the software framework in orderto provide a software environment for algorithm;(4) The system parameters were calibrated and a measurement experiment wasaccomplished to verify the feasibility of the program and the results was analyzed.The analysis of the experimental results shows that the measurement system withcomputer vision could measure the pitch of transmitting signal plate fast and non-contact, which can accelerate the speed of production while strengthening thequality monitoring and has great significance to promote the development of Autoindustry.