Research On The Enhancement Methods Of Integral Imaging Spatial Resolution

Integral Imaging (II) is a three-dimensional (3-D) imaging technique including thecapability of full-parallax, multi-viewing, working with incoherent light source, and isviewable without the requirement of special devices. As a result, it has been regardedas one of the most promising techniques for3-D imaging and display. However, thereconstructed images of an II system have some problems, such as the low resolution,etc. In this paper, the whole process of computational generation and reconstruction ofII is realized, and focusing on a practical II system, several methods to improve theimage resolution and quality of II are studied.Firstly, based on the principle of the Virtually Curved Integral Imaging (VCII)system, the optical ABCD matrix theory is utilized to derive the mapping relationshipbetween the three-dimensional object and3D integral image, the VCII computationalgeneration and reconstructed image is realized. And the advantage of VCII inimproving the integral image quality through increasing the sampling rate duringpick-up process is confirmed.Secondly, incorporating the principle of micro-scanning and II, the characteristicsof II based on micro-scanning method is analyzed. The equation of depth resolutionand the quantitative relationship between the depth of focus and lateral resolution of IIare derived. Then based on micro-scanning, the II process is simulated computationally,the quality of computational reconstructed images within various modes ofmicro-scanning is analyzed.Finally, focusing on the problem of the existence of pixels without information inthe images reconstructed by the Direct Point-to-point Mapping method, twointerpolation reconstruction methods are analyzed, by which, the image qualities ofreconstructed images are enhanced. In addition, this paper analyzes the enhancementeffects of the reconstructed images resolution, with the combination of themicro-scanning method in the pickup process and the proposed bi-directioninterpolation algorithm in the reconstruction method.