Detection Of Diamond Particle Based On Image Analysis

With the development of the digital image processing technology, the detection technique of particles plays a more and more important role in production detection of industrial fields.The thesis introduced a series of diamond particle detections and analysis methods, especially detections of different kinds of diamond particles. The software is programmed with LabVIEW (Laboratory Virtual instrument Engineering) , a graphical language of program. The work includes following aspects:Firstly, the segment methods of a particle or two particles in an image are presented. Because diamond particle has small ranges in hue value and high luminance, after converting the original image from RGB color space to HSI(Hue, S Saturation, Intensity) color space, applying different thresholds to three planes of an HSL image, the original image become binary image.Secondly, some diamond particle analysis methods are introduced to detect different kinds of parameters in one image. With the above result, the area and perimeter parameters are measured. Then, the circularity factor, ellipse factor and particle size are calculated according to distance of parallel edges. If there is a split in the diamond particle image then it needs to repair.Thirdly, the quantitative analysis and location of diamond particles in the original image is implemented. After preprocessing including converting color image to gray-level image, image enhancement, and binary image analyzing, thelocation of the particle's centroid and the number of the particles are calculated. The reports of the results are given.Finally, the experimental results of diamond particles detection show that the measurement become more efficient and supply more accurate parameters of diamond particles sample. At the same time, it will reduce the artificial errors. Furthermore, the soft can be used in corresponding image analysis fields.