| Vision is an important way for humans to obtain outside information.The real objective world is three-dimensional(3D).However,traditional two-dimensional(2D)images always lack depth information,which cannot satisfy humans' visual demand to reconstruct the 3D world.In recent years,3D technology has received widespread attention due to its demand and application.Among 3D display techniques,integral imaging is considered as one of the most promising,because it doesn't need to wear auxiliary viewing equipment,works under incoherent light sources,and has the advantages of full-color parallax,continuous viewing angle,and no visual fatigue.Since it was proposed by Lippmann in 1908,the integral imaging 3D display system has developed to a certain extent.However,due to the limitations of theoretical algorithms and hardware equipment,integral imaging still needs to be improved in 3D scene acquisiton,computational reconstruction,and optical display.In terms of acquiqition,the camera array used in integral imaging technology has excessive hardware costs.In terms of computational reconstruction,the traditional method has image distortion.In terms of optical display,traditional display systems have problems such as serious light crosstalk and strong graininess.Therefore,after analysis of related theories,this thesis focuses on the key issues such as 3D scene acquisition and processing,computational integral imaging reconstruction,and the visual quality improvement of integral imaging display system.The deficiencies of the current algorithms and display systems are analyzed,then improved solutions are proposed.The effectiveness and feasibility of the proposed methods are proved by theoretical analysis and experimental data.The main research contents and innovative work of this thesis are as follows:1.Acquisition method based on scale invariant feature transform(SIFT)and discrete viewpoint images in integral imaging.In order to obtain high-quality real-scene display content,camera arrays are generally used for acquisition in integral imaging,but using the same number of cameras as the display lens will cause high costs.Hence for capturing high-quality real scenes while saving cost,an integral imaging acquisition method based on discrete viewpoint imagings and SIFT descriptors is proposed.This method uses a sparse camera array to capture discrete viewpoint images and small-scale elemental image array,and then converts the small-scale elemental image array into a low-resolution sub-image array.The discrete viewpoint images have enough effective pixels to form the large-scale elemental image array for display,but they lack accurate perspective information.In contrast,the low-resolution sub-image array has accurate perspective information,but there are not enough effective pixels in it.Due to the similarity between the discrete viewpoint images and the low-resolution sub-images,SIFT descriptors are used to match these two sets of images,and then the discrete viewpoint images are transformed into high-resolution sub-images according to the matching results.Finally,the high-resolution sub-image array is converted into a large-scale elemental image array that meets the display requirements.This section not only performs virtual simulation but also actual experiments on the acquisition process and display results,which proves the effectiveness and feasibility of the proposed method.This method can be used as a theoretical basis for the design of integral imaging acquisition system.Compared with the computational virtual acquisition method,the proposed method can effectively acquire real scenes.Compared with the acquisition method based on lens array and recording medium,the proposed method can flexibly acquire scenes of various size.Compared with the traditional camera array acquisition method,the proposed method can effectively reduce the acquisition cost,greatly reduce the complexity of the acquisition operation,and make the generation of the elemental image array more convenient and efficient.2.3D view image reconstruction in computational integral imaging using SIFT and patch matching.Aiming at the problem of local image distortion in the 3D view image directly generated from the elemental image array,a method of 3D view reconstruction based on SIFT and patch matching is proposed.This method uses the correct regions obtained from the view images taken directly from the original object or use patch matching to replace the distorted regions.In fact,due to the limitation of equipment and inevitable shortcomings of the experimental operation,the initial matching regions cannot meet the requirements of the replacement process,so the SIFT descriptor and perspective transformation are used to obtain the correct regions.In this section,the experimental results of virtual acquisition and real acquisition are given to verify the proposed method,and the image quality evaluation proves that the proposed method can improve the visual quality of the 3D view images.Compared with the traditional computational integral imaging reconstruction method,the 3D view image generated by the proposed method has a higher resolution.Compared with the computational integral imaging reconstruction method based on ray tracing and auto-focus,the proposed method not only reduces the computational burden and solves the problem of blurred edges in reconstructed scenes.3.Integral imaging display method based on discrete glued lens array and holographic diffuser.To solve the problems of traditional integral imaging,such as poor visual quality,the ray-crosstalk in display,and obvious graininess,a display method based on discrete glued lens array and holographic diffuser is proposed.The proposed method replaces the continuous single lens array in the tratidional integral imaging display method with a discrete glued lens array and a holographic diffuser.In this part,the structure and imaging quality of single lens and glued lens are designed,analyzed and compared.And the diffusion effect of the holographic diffuser is theoretically analyzed.We have designed two display system based on the ultra-high-densty small-pitch LED display plane.The experimental results show that the visual quality of the proposed method is significantly improved compared with the traditional integral imaging display method.Compared with the traditional integral imaging display method based on a continuous single lens array,the proposed method can effectively reduce the influence of ray-crosstalk on the 3D images,smooth the discontinuous light field distribution,and reduce graininess to improve visual quality.In addition,since the traditional continuous single lens array needs to manufacture a mold tool with the same size as the display platform during processing,and the discrete glued lens array only needs to process the unit lens and then assemble them.If a large-scale integral imaging display system is to be manufactured,discrete lense array is more suitable,because the discrete glued lens array is cheaper and easier to be manufactured.In order to meet the experimental expectations and improve the display visual quality,we used a glued lens array in the experiment.Compared with aspheric lens that is difficult to process,the glued lens is easier to design and process and has good display visual quality. |
PDF Full Text Request