| Self-interference digital holography is an attractive technology as it dispenses with the dependence on coherent light source and possesses the advantages of speckle-free noise and non-scanning three-dimensional imaging. In the process of recording incoherent holograms, each object point corresponds to a hologram resulting from the interference of objective beam and self-informative reference in the recording plane, and the superimposing of all the holograms form the complex hologram of the original object. Self-interference digital holography can be used to realize recording spectrum information of objects, color imaging and multi-wavelength fluorescence imaging by making full use of the wavelength information of light source. However, self-reference digital holography is not a technology of three-dimensional tomographic imaging because the reconstruction of 3D information from a two-dimensional(2D) hologram is ill-posed. Compressive sensing is a kind of model allows to reconstruct the original data from fewer measurements than the number determined by Shannon's sampling theorem. Compressive holography is a technology combines digital holography with compressive sensing. In this paper, self-interference compressive digital holography is developed according to the investigation of compressive Fresnel digital holography. The basic principle of this technique, reconstruction condition, optimization algorithm, and the application of non-scanning and multi-dimensional imaging are investigated. We study the theoretical model of compressed digital holography, build the forward sensor model under self-interference holographic recording scheme, and demonstrated the reconstruction conditions. Based on this theoretical model, the problem of reconstructing multi-dimension original data from low dimensional measurements is solved. The main researches are summarized as the following: 1. The basic principle of the compressive digital holography is investigated. Based on compressive sensing theory, we regards the Fresnel propagation process as a sensing operator, and holographic recording is a process of compressing muti-dimensional data into a two dimensional hologram. Thus the compressive holographic recording model is established. With this model, simulations and experiments are completed. The object information is completely reconstructed from sub-sampling hologram2. The effect of sparsity and incoherence on the reconstruction in compressive Fresnel holography is investigated. The dependence of reconstructed imaging quality on different physical properties, sampling methods and system noise are demonstrated. From the above investigation, the optimized combination of reconstruction conditions is specified for improving the reconstruction quality. All these investigation provide the theoretical basis of multi-dimensional imaging. 3. Compressive sensing and self-interference digital holography is combined in order to improve and extend the imaging performance of self-interference digital holography. Self-interference digital holography can be regarded as a kind of sensing operator and establish the forward sensing model of self-interference digital holographic recording. Specifying the relationship between parameters during recording. The self-interference digital holography was verified obeying well compressive sensing framework. Both theoretical analysis and experimental results are demonstrated. 4. The application of self-interference compressive holography in multi-dimensional imaging is investigated. The object of each wavelength channel is captured by a color imaging sensor using the self-interference holographic system. The object data with wavelength information and three dimensional(3D) information is reconstructed by the compressive sensing algorithm. The image quality of CS reconstruction is better than back propagation reconstruction. A new type technology of a non-scanning and multi-dimensional imaging technology is developed. |
PDF Full Text Request