Image Quality Evaluation of Light Field Photography

Light field photography captures 4D radiance information of a scene. Digital refocusing and digital correction of aberrations could be done after the photograph is taken. However, capturing 4D light field is costly and tradeoffs between different image quality metrics should be made and evaluated. This paper explores the effects of light field photography on image quality by quantitatively evaluating some basic criteria for an imaging system. A simulation approach was first developed by ray-tracing a designed light field camera. A standard testing chart followed by ISO 12233 was provided as the input scene. A sequence of light field raw images were acquired and processed by light field rendering methods afterwards. Through-focus visual resolution and MTF were calculated and analyzed. As a comparison, the same tests were taken for the same main lens system as the results of conventional photography. An experimental light field system was built up and its performance was tested. This work helps better understanding the pros and cons of light field photography in contrast with conventional imaging methods and perceiving the way to optimize the joint digital-optical design of the system.