Implementing light field image refocusing algorithm

• Published in: 2015 2nd International Conference on Opto-Electronics and Applied Optics (IEM OPTRONIX) pp. 1 - 8
• Main Authors: Wenxing Fu, Xi Tong, Chuan Shan, Supeng Zhu, Baijia Chen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.10.2015
• Subjects: Algorithm design and analysis; algorithms; Arrays; Cameras; computational frequency; Focusing; Fourier slicing theorum; frequency domain refocusing; Frequency-domain analysis; image processing; Lenses; light field imaging; space domain refocusing
• DOI: 10.1109/OPTRONIX.2015.7345517

The light field imaging produces the information on the direction of light ray, therefore the refocusing of the raw image shot with the light field camera can acquire high-resolution images of objects at different distances. This paper presents the light field image refocusing principles, rearranges raw images, transforms virtual image planes and integrates light rays, thus obtaining ideal refocused images. Then it implements frequency domain refocusing and space domain refocusing and applies the Fourier slicing theorem to implementing the frequency domain refocusing algorithm and uses the fast Fourier transform to enhance its efficiency. Through experiments, it compares the computational efficiency of the space domain refocusing algorithm and the frequency domain refocusing algorithm. It concludes that the latter has a higher computational efficiency.