Plenoptic flow: Closed-form visual odometry for light field cameras

Three closed-form solutions are proposed for six degree of freedom (6-DOF) visual odometry for light field cameras. The first approach breaks the problem into geometrically driven sub-problems with solutions adaptable to specific applications, while the second generalizes methods from optical flow to yield a more direct approach. The third solution integrates elements into a remarkably simple equation of plenoptic flow which is directly solved to estimate the camera's motion. The proposed methods avoid feature extraction, operating instead on all measured pixels, and are therefore robust to noise. The solutions are closed-form, computationally efficient, and operate in constant time regardless of scene complexity, making them suitable for real-time robotics applications. Results are shown for a simulated underwater survey scenario, and real-world results demonstrate good performance for a three-camera array, outperforming a state-of-the-art stereo feature-tracking approach.

I. Introduction

Visual odometry has a long standing as a fundamental and useful application of computer vision. Due to the breadth of applications, low cost of camera'¼s, and wealth of information presented by the visual world, the deceptively simple-sounding problem of tracking a cameras trajectory continues to attract attention [1], [2]. Meanwhile, decreasing camera costs have made camera arrays a practical alternative to monocular and stereo rigs, with recent advances demonstrated in the closely related fields of generalized cameras, light field or plenoptic cameras, and polydioptric cameras [3]–[5].