Algorithm Study On Dimension Measurement Based On Image

Product quality and production efficiency must be impacted directly by the accuracy and rapidity of measurement of component dimension. With the rapid development of modern manufacturing, the needs of high-precision and high-speed measurement of component dimension can’t be satisfied absolutely by traditional means of measurement in the field of industrial production. Therefore, it is so essential and significative to study how to improve the accuracy and rapidity of measurement of component dimension. Dimension measurement system based on image was built to study how to measure component dimension accurately and speedily by means of digital image processing techniques in this paper.Further research on image measurement technology of component dimension was done in this paper. Firstly, the conversion of BMP image, from32bit ARGB color mode to24bit RGB color mode, was researched, which can reduce availably nonessential image data; Mean filter and median filter occur in the form of different filter templates, and filtering effect of different filter templates was compared, Composite filter can effectively reduce hybrid noise and SNR is improved; The improved gradient sharpening highlights the image edges, but also retains the original appearance of gentle change area and enhancement effect and applicability is better than direct gradient sharpening; Image preprocessing techniques must be capable of suppressing noise and enhancing contrast in some degree, which can improve significantly the accuracy of the image segmentation and shape extraction; Then a comprehensive comparison study on image segmentation algorithm based on the threshold was conducted in this paper. With genetic algorithm, global optimal solution for threshold segmentation was achieved; Research on curve fitting of coordinates of image edge point was also done on the basis of segmentation. Meanwhile with CCD camera, PCI video card and PC, image measurement system of component dimension was built, and with VC6.0integrated development environment, measurement interface was designed and program of image processing algorithm was written, through the system calibration, pixel equivalent was determined, and parameter of rectangle and circle component was measured via measurement system; Finally, the source of error of measurement system was analyzed, and the measures were given to improve the measurement precision.