Research On 3D Integral Imaging Visualization Under Limited Conditions

Three dimensional(3D)integral imaging(II)system is one kind of 3D stereo imaging techniques with various merits such as continuous viewing points,full parallax,no visual fatigue and simple imaging structure.However,due to the property limitation of imaging sensor,the II system often suffers poor visual quality of 3D reconstruction under limited conditions,such as multi-focus and low irradiance.In this paper,visualization enhancement solutions of 3D integral imaging have been proposed to solve these problems.The limited depth of field(DoF)of image sensor in II system will cause the multi-focus problem,which means the objects in the DoF are focused while others out of the range will turn into be blurred.The situation could heavily degrade the viewing quality of 3D reconstructed images.In this paper,a three dimensional depth of field enhancement system in integral imaging has been proposed by considering both the theory of 2D image fusion techniques and the 3D II system imaging structure.Firstly,foreground object images in elemental images are extracted and then feature information of image clarity is recorded from each object image and made a comparison to select the clearest one as the final mask in image fusion.Finally,a set of 3D slice images are reconstructed with the fused elemental images.The preliminary experiments are performed and certify the proposed system can preserve the perspective information of object and solve the defocus problem caused by multi-focus as well,which can generate a set of clear elemental images and improve the DoF of II system.Under the limited irradiance condition such as at the dark night without moon or under the deep sea,the picture captured by conventional optical image sensor will lost a huge amount of feature information and heavily impacts the visual quality of 3D reconstruction.In this paper,a three dimensional photon counting integral imaging(PCII)system is utilized to solve the issue.Firstly,a set of photon counting elemental images is captured and then statistical approaches are utilized to optimize the original photon counting images,finally 3D reconstruction images are obtained through II system.Considering the 3D stereo imaging structure of II system,we optimize the conventional statistical estimation algorithms to estimate the missing photons in photon counting images more precisely.The preliminary experiments verify the efficiency of our method to improve the 3D visualization of integral imaging under low irradiance condition.What's more,the application of PCII system has been extended in this paper based on the principle and property of the system.For example,during the 3D visualization under high scattering media condition,a high quality reconstruction image can be obtained based on PCII system.And the sparse property of PCII system is just corresponding with compressive sensing technique,which can further improve the compressive ratio,speed up the transmission rate,and recover the transmitted image efficiently.