Research On Methods Of Particle Size Measurement Based On Technology Of Digital Holography

With the development of science and technology in electronics, chemicals,machinery, metallurgy, pharmaceuticals and environmental protection, the accuratemeasurement of the particle size and shape is needed. The particle size measurement isvery important to the industrial and agricultural production, the environmental protection,national defense and human health significance. In many particle size measurement areas,there are many measurement methods to choice. In recent years, with the rapiddevelopment of the high-resolution CCD and computer technology, the digital in-lineholographic particle size measurement method has become a hot research focus becauseit can realize dynamic measurement and get details information of the three-dimensionalparticle field. This paper carries out research about the particle size measurement in threeaspects. It is the recording and reconstruction of particle field digital in-line holographic,the quality improving of reconstruction image and the analysis of particle fieldreconstruction image. The main research works of the thesis are as follows:First of all, based on the optical diffraction theory of light, the mathematicaldescription of digital holographic recording and reconstruction is analyzed. And thedigital holographic reconstruction algorithm of the Fresnel transform method is comparedwith the convolution method. The wavelet transform reconstruction algorithm of digitalholographic is detailed derived, and the simulation experiments and the comparing aredone to the convolution algorithm and the wavelet transform algorithm. The digitalin-line holographic recording system theory of plane wave and spherical wave of theparticle field are analyzed. And the particle field measurement optical system andmeasurement model are established.Secondly, in order to improve the particle field digital holography reconstructionimage quality, the theoretical basis of the wavelet threshold denoising is studied, anddigital holography speckle noise model is established. And two specific speckle noiseremoval algorithms based on the ideas of edge detection and wavelet transformthresholding noise removal are put forward. In the first method, the wavelet modulus maxima edge detection algorithm is improved, and the adaptive threshold based on a dualthreshold algorithm is proposed to achieve more accurate edge detection. The improvedalgorithm and wavelet threshold denoising based on the Neyman-Pearson criterion areapplied to suppress the speckle noise of digital holographic reconstruction images. In thesecond method, the wavelet threshold function is improved. A compromise thresholdfunction is proposed, and the threshold function can both ensure the smoothness of thereconstructed image, and better highlight the local features of the image. And theimproved compromise threshold function is applied to digital holographic reconstructedimages speckle noise suppression. Experimental results show that the two methods cansuppress digital holography speckle noise and keep better the edge of the imagesimultaneously, and the reconstructed image quality is improved.Finally, the hologram reconstruction image analysis method is used to measure theparticle size of the particle field. The point spread function of the digital the in-lineholographic imaging system is systematical studied, and the relationship between theparticle image diameter and the true diameter is theoretical analyzed. A set of digitalimage processing method is proposed to obtain the particle spatial location and particlesize from the reconstructed images. The two key issues of particle extraction and particlepositioning are studied, and the edge detection method is used to separating the particlesfrom the background. And a particle positioning algorithm based on wavelet transformmodulus average value is proposed, the positioning algorithm distinguishes particle focusplane in the transform domain and can achieve precise particle positioning. The chaincode technology and regional target area calculation method are studied, and the Freeman8-chain code target area calculation method is used to calculate the particle size toimprove the efficiency and accuracy of particle size computing. Finally, the imageprocessing method is used in the experimental measurement of the standard particles, andthe experimental results demonstrate the feasibility of the image analysis and processingmethod.