Research On Digital Holographic Differential Autofocusing Technology

Digital holographic microscopy(DHM)has the advantages of high resolution,full field,real-time,and non-invasiveness,and has become one of the indispensable tools in modern scientific research.At the same time,DHM can realize numerical refocusing inside the computer,identify the focal plane of the target image by defining the refocusing criterion,and then provide a reliable and clear image for quantitative imaging analysis of the object.The traditional auto-focusing criterion determines the focal plane by judging the peak or valley value of the focusing evaluation function,but in the precise self-focusing process,the discrete step size is small,the sensitivity near the peak/valley is low,and the focusing accuracy is easily affected by noise.Increased difficulty in determining the best focus position.Subject "Research on digital holographic differential Autofocusing Technology" is studied on the structure of single-wavelength and dual-wavelength digital holography.On the basis of the traditional self-focusing algorithm,by introducing a virtual differential optical path,the accurate reproduction distance is obtained and the curve is improved.Noise immunity.The specific research contents are as follows:Firstly,a autofocusing algorithm based on spectral fractional difference is proposed for the single-wavelength digital holographic structure.The algorithm considers that the adjacent reconstructed spectra are closer in the focal region than in the defocused region,and uses the spectral score function to evaluate the correlation between the adjacent spectra of the two axes.Greatly improve operational efficiency.On this basis,a virtual differential optical path is introduced,and two virtual detectors are placed on both sides of the real detector.This structure converts the traditional peak-valley judgment method into zero-crossing judgment,avoiding the low peak-to-valley sensitivity of the traditional self-focusing algorithm.Secondly,a autofocusing algorithm based on complex field similarity difference is proposed for the dual-wavelength digital holographic structure.The algorithm obtains the reconstructed light fields of two wavelengths at different defocus distances,and uses the complex field similarity function to evaluate the cosine similarity between the reconstructed complex amplitudes at different wavelengths.Then,a virtual differential optical path is introduced to realize dual-wavelength differential motion,and the focal plane is positioned at the zero-crossing point of the differential curve to avoid the problem that the peak-to-valley value of dual-wavelength self-focusing is susceptible to interference.Dual-wavelength selffocusing needs to establish multiple series of reconstructed images.The proposed algorithm is combined with spectrum sampling technology to reduce the number of pixels in the focusing iteration process by reducing the size of the object light diffraction field,and improve the speed of dual-wavelength autofocusing.Finally,a autofocusing software is designed for the proposed algorithm.This software encapsulates the algorithm proposed in Chapter 3 and integrates image input,processing and output.Users can select the image to be processed according to their needs,and finally obtain the focused image directly by the algorithm in the software.The software is simple and easy to operate.