Inertial-aided KLT feature tracking for a moving camera

We propose a novel inertial-aided KLT feature tracking method robust to camera ego-motions. The conventional KLT uses images only and its working condition is inherently limited to small appearance change between images. When big optical flows are induced by a camera-ego motion, an inertial sensor attached to the camera can provide a good prediction to preserve the tracking performance. We use a low-grade MEMS-based gyroscope to refine an initial condition of the nonlinear optimization in the KLT. It increases the possibility for warping parameters to be in the convergence region of the KLT. For longer tracking with less drift, we use the affine photometric model and it can effectively deal with camera rolling and outdoor illumination change. Extra computational cost caused by this higher-order motion model is alleviated by restraining the Hessian update and GPU acceleration. Experimental results are provided for both indoor and outdoor scenes and GPU implementation issues are discussed.

I. INTRODUCTION

Feature tracking is a front-end to many vision applications from optical flow to object tracking and 3D reconstruction. Higher-level computer vision algorithms require, therefore, robust tracking performance no matter how a camera moves. KLT (Kanade-Lucas-Tomasi) feature tracking which uses template image alignment techniques has been extensively studied from the seminal work by Lucas and Kanade [1], Shi and Tomasi [2], to the unifying work by Baker and Matthews [3].