Research On Micro Parts Dimention Measurement System Based On FPGA

With the developments of modern science and technology, micro instrumentdevice is more and more intelligent, miniaturized, sophisticated, and micro instrumentdevice is applied to more and more widely the demand of the micro parts also willincrease. The quality of micro parts is bound to affect the features of the miniaturemechanical devices, quality inspection of micro parts is particularly important. Thesize detection of the micro parts is one of important test items. The parts detectionsystem based on machine vision technology is divided into machine vision detectionsystem based on computer and machine vision detection system based on embeddedprocessor. This paper analyzes the deficiency of the machine vision detection systembased on computer, and proposes a machine vision detection system based on a singlechip of FPGA embedded processor. The system can well satisfy the measurement ofmicro parts.At first, this paper builds general structure of the machine vision detection systemof micro parts size based on FPGA. According to the corresponding functional, themachine vision system is divided into mechanical parts, optical parts as well as theimage acquisition and processing based on FPGA. On mechanical parts of the system,this paper mainly completed the selection of model of mechanical part, including theselection of model of3-D measurement servo motion mechanism and measurementplatform of system.This paper completed the design of hardware architecture and related software ofimage acquisition and processing system based on FPGA. The design of hardware partincludes the selection of the main chip and circuit design, including the selection ofFPGA chip, storage, video decoder, video coder and the image display device, then thecircuit design. The design of software of the image acquisition and processing systemanalyses the related design software. including the design process of QuartusⅡ,MATLAB/Simulink, DSP Builder. This paper analyzes how to build the correspondingmodule about the SOPC system and build a machine vision detection system based ona single chip FPGA.In this paper, after the size detection system of micro parts was constructed, thispaper carried out the size detection of the micro parts. In machine vision system of sizedetection of parts, the size detection of micro parts was analysis of the image processing of micro parts. Image processing generally included image acquisition,grizzled processing, smoothing filtering, binarization, edge detection, edge fitting, sizecalculation. In the conversion process of the size of parts from pixel size to actual size,system calibration was needed. The relationship between actual size and pixel size wasdetermined by system calibration，so the value of the actual size of the parts measuredwas obtained. Finally，this paper selected micro gears as object to be measured，andcompleted the preprocessing of the image of micro gears. The preprocessing of theimage of micro gears included grizzled processing, smoothing filtering, binarization.After the preprocessing of the image, edge detection of the image of micro gears wascompleted. This paper used the method of least squares to fit the edge of the image ofmicro gears. The size of micro parts was obtained after calibrating the system.