Abstract
This work proposes the use and analyzes the viability of graph Fourier transform (GFT) for light-field compression. GFT is employed in place of discrete-cosine transform (DCT) in a simplified compression system based on high-efficiency video coding (HEVC). The effect on GFT efficiency of different implementations for prediction procedure is analyzed, as well as different methods for computing GFT given residual images. Results indicate that the prediction scheme is sensitive to the type of light field being compressed, and a preliminary method for selecting the best prediction scheme is explored. Moreover, considering multiple residual images when computing GFT, instead of only one central image, improves compression rate and makes compression more uniform across multiple views. GFT achieves reduction of up to 21.92% in number of transform coefficients when compared to DCT-based compression, while providing better or equal mean squared reconstruction error.
Highlights
Light field imaging is a promising technology that opens a variety of new possibilities to entertainment industries, such as photography and cinema, by capturing 4D data from a scene [1]–[7]
The basic concept underlying all simulations presented in the subsections is to set graph Fourier transform (GFT) coefficients to zero as much as possible while still recovering blocks with less distortion when compared to a specific discretecosine transform (DCT) compression
The number of compressed DCT coefficients is fixed at QDCT = 924, i.e., only the 100 largest out of 1024 coefficients are kept and DCT is fixed at approximately 10:1 compression ratio
Summary
Light field imaging is a promising technology that opens a variety of new possibilities to entertainment industries, such as photography and cinema, by capturing 4D data from a scene [1]–[7]. If the plenoptic function for a region of interest is known, any image associated with that region can be recreated, from every perspective This motivates the use of light field in entertainment industries, mainly photography and cinema [1]. Determining the plenoptic function is unfeasible, so light field cameras capture a 4D parametrization of the plenoptic function that consists of multiple photographs of a scene. This can be done moving a digital camera in a grid of various positions and taking photographs at each position, by using an array with multiple cameras, or by adding a microlens array in front of the camera sensor [3]
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
