Abstract
Multi-view or light field images have recently gained much attraction from academic and commercial fields to create breakthroughs that go beyond simple video-watching experiences. Immersive virtual reality is an important example. High image quality is essential in systems with a near-eye display device. The compression efficiency is also critical because a large amount of multi-view data needs to be stored and transferred. However, noise can be easily generated during image capturing, and these noisy images severely deteriorate both the quality of experience and the compression efficiency. Therefore, denoising is a prerequisite to produce multi-view-based image contents. In this paper, the structural characteristics of linear multi-view images are fully utilized to increase the denoising speed and performance as well as to improve the compression efficiency. Assuming the sequential processes of denoising and compression, multi-view geometry-based denoising is performed keeping the temporal correlation among views. Experimental results show the proposed scheme significantly improves the compression efficiency of denoised views up to 76.05%, maintaining good denoising quality compared to the popular conventional denoise algorithms.
Highlights
Multi-view images or light field data, acquired from a multi-camera array by capturing a plurality of images of different viewpoints, are used in various applications, including the synthesis of a high-resolution image with a high dynamic range [1] or the reconstruction of an occluded area through synthetic aperture photography (SAP) [2]
Considering v2 is located to the right of v1, epipolar line searching (ELS) is performed toward the left of the reference pixel r2 in the range of the SR1step which is set to the maximum disparity that can be varied depending on the distance between the camera and the closest object in multi-views
The proposed scheme in this paper improves the consistency of denoised multi-view as well as the performance of noise reduction by combining an excellent denoising algorithm block matching 3D collaborative filtering (BM3D) with
Summary
Multi-view images or light field data, acquired from a multi-camera array by capturing a plurality of images of different viewpoints, are used in various applications, including the synthesis of a high-resolution image with a high dynamic range [1] or the reconstruction of an occluded area through synthetic aperture photography (SAP) [2]. Camera array is used, a highly accurate depth map can be obtained from the 3D focus image stack It does not work well for views with severe noise or wide view distances. Multi-view image denoising algorithm based on convolutional neural network (MVCNN) [22] that represents an enhancement of earlier work was proposed [21]. Proposed the EPI-based depth estimation that utilized a spinning parallelogram operator (SPO) for light field camera images with noise and occlusion [30]. The 3D filtering and the multi-view geometry-based EPI estimation are effectively combined to speed up the multi-view denoising.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
