压缩光场成像：原理、方法及应用

打开文本图片集

中图分类号：TN278 文献标志码：A

Abstract： Light field imaging records both spatial and angular information of light rays in a 3D scene， enabling capabilities such as viewpoint variation and depth perception that are not possible with traditional 2D imaging. As computational photography and imaging technologies advance， light field cameras have evolved from conceptual designs to practical applications through various acquisition techniques. This review systematically examined the development history and fundamental concepts of light field imaging， including the plenoptic function and four-dimensional parameterization of light fields. We summarized diferent implementation approaches for light field data acquisition and their representative research， with particular emphasis on the basic principles and current state of compressive light field imaging. Finally， the article summarized the typical applications of light field imaging in various fields， such as 3D imaging，augmented reality/mixed reality （AR/MR）， and industrial inspection. However， despite numerous advancements，light field imaging still faces challenges in data storage， realtime processing， and hardware complexity. Looking ahead， with continuous innovation in optical encoding technology and optimization of deep learning algorithms， compressive light field imaging is expected to achieve broader applications in multiple fields.

Keywords： light field imaging; compressive light field imaging; camera array; micro-lens array; deep learning

光场成像的发展历程与基础概念

光场成像技术被认为是传统应用光学的一个革命性的技术，是当下最有前景的三维成像技术之一。（剩余31387字）