Robust Image Alignment for Outdoor Teach-and-Repeat Navigation

Visual Teach-and-Repeat robot navigation suffers from environmental changes over time, and it struggles in real-world long-term deployments. We propose a robust robot bearing correction method based on traditional principles aided by exploiting the abstraction from higher layers of widely available pre-trained Convolutional Neural Networks (CNNs). Our method applies a two-dimensional Discrete Fast Fourier Transform based approach over several different convolution filters from higher levels of a CNN to robustly estimate the alignment between two corresponding images. The method also estimates its uncertainty, which is essential for the navigation system to decide how much it can trust the bearing correction. We show that our "learning-free" method is comparable with the state-of-the-art methods when the environmental conditions are changed only slightly, but it out-performs them at night.