Research On Autofocusing And Image Enhancement Technology In Digital Holography

The paper aims to study the key technology of digital holography. Through the method of combining theoretical analysis and experimental research, on the basis of domestic and foreign digital holography research conducted in-depth research and analysis, the paper studies three key technologies of digital holography which are as follows:(1)To simplify the large object lensless Fourier transform digital holographic experiments set, the paper studies a method applying to large object lensless synthetic aperture digital holographic imaging. Through modifying the Mach-Zehnder interferometer, this method splits the raw reference beam into two branches and redirects one of them as a second object beam. The two object beams will illuminate different sections of a large object and interfere with the reference beam creating a hologram on a CCD. Then, using synthetic aperture techniques, the paper can achieve a digital holographic reconstructed image under single-exposure.(2)In the process of digital holographic reconstruction autofocusing, in order to reduce the influence of the mean value drifting of reconstructed images on several criterion functions, the paper proposes a new strong adaptable autofocusing criterion function named as the standard deviation contrast of the high-frequency component method. This method is calculating the criterion with only the original hologram’s high-frequency components and reducing the mean value drifting problem existing in the conventional methods. It is more simple and efficient. In the digital hologram reconstruction of different image qualities, it shows strong adaptability.(3)In the process of digital holographic reconstruction, the reconstructed image is degraded by speckle noise. In order to improve the image quality of digital holographic reconstructed image, the paper raises an efficient algorithm for image quality enhancement named as the enhancement of low quality reconstructed image of digital hologram based on frequency extrapolation of large object under diffraction limit. This approach first lower the speckle noise by an adaptive Gaussian filter from the Fourier filtered image reconstructed from the +1 order of the original hologram. It is known as the original spectrum. The predicted spectrum will be acquired through the continuous gradient analytical iteration. Finally, the reconstructed spectrum with predicted high frequency information of the restoration image will be achieved by the synthetic correction of the original spectrum and the predicted spectrum.