Scalable coding of 3D holoscopic image by using a sparse interlaced view image set and disparity map

3D holoscopic imaging, also known as integral imaging, light field imaging or plenoptic imaging, can provide natural and fatigue-free 3D visualization. Holoscopic contents captured by the plenoptic camera contain both spatial and angular information of a 3D scene. Therefore, view images with different perspectives can be rendered from the holoscopic contents. A coding scheme to compress the 3D holoscopic image by exploiting the high spatial correlation among the rendered view images will be advantageous. Therefore, in this paper, an efficient scalable coding scheme is proposed to compress the 3D holoscopic image by utilizing such high spatial correlation. We firstly re-arrange the holoscopic contents to form an interlaced view image. A sparse format is then proposed to express the interlaced view image. With the reconstructed image derived by disparity map based sifting and interpolation, the full interlaced view image is coded by using the reconstructed image as a reference frame. As an outcome of the representation, a scalable structure with three layers can be provided by the proposed scheme. Experimental results demonstrate that the 3D holoscopic image can be compressed efficiently with over 44 percent bit rate reduction compared with HEVC. Meanwhile, the proposed scheme can also surpass several other prediction schemes in this field.