Research On The Key Technology Of Diamond Particle Morphology Detection System

Diamond particles mainly rely on men using microscope to measure the characteristic parameters for assessment the grade in our country,which has a large amount of defects.In addition, with the mature application of digital image processing technology in various industries, the recognition and measurement method based on machine vision has become an important development direction of the quality control of diamond particles.The domestic research of particle detection technology based on machine vision started relatively late,foreign advanced detection instrument is more expensive and the related technologies are confidential.Domestic related research most involving single elements, therefore the study of the key technology of diamond particle morphology detection, which has a very important significance to identify the grade of diamond particles rapidly.This paper further studies the key technology of diamond particle morphology detection system,the main content is as follows:According to the characteristics of small particle size, highly reflective surface and multiple detection parameters,this paper design a diamond particle morphology structure of the detecting system, including the hardware platform and the software design in machine vision module for the detection system. The image acquisition system composed of CCD industrial camera, microscope, light source and so on was constructed.In view of the reflection of the surface of the diamond particles, the LED positive unilateral illumination mode is used to get a better image effect.This paper study the method of measuring parameters of diamond particles based on image processing technology. First to filter the existing noise in the acquisition process because of the light, shoot angle, signal transmission, discusses some common methods of the spatial filtering and edge detection algorithm, and study the canny edge detection algorithm based on median filter to extract the better particle edge. For the extraction of the non closed particle edge, this paper put forward contour edge location method based on Hough transform, according to the polar radius and polar angle of the boundary line, realize the precise location and fitting of the contour line and get the closed polygon edges of diamond particles. Finally, according to the perimeter and area of and other parameters, the parameters such as particle size, ellipse and roundness of diamond particles are measured.The color feature is an important index to measure the quality of diamond particles. According to the color composition of diamond crystal, This paper propose a model of color image segmentation method based on the binarization method, translate the RGB color model of image into HSV color space according with human visual perception and combine the H, S color parameters for feature extraction to achieve the measurement of color. Based on the analysis of the characteristics of the gray scale image and the flaw and yellow area bimodal histogram feature of the particle, the Otsu method was used to identify and select the flaws.VC++,OpenCV and SQLite are selected to build the software platform of image processing system.This paper carry out measurement for series of diamond particle samples, and achieve diamond particle shape, size, clarity, penetration and other parameters. Through the error analysis of experimental results of the measurement system, the method validity, reliability and feasibility. are verified.