Edge Enhanced Direct Visual Odometry

In this paper, we propose an RGB-D visual odometry method that both minimizes the photometric error and aligns the edges between frames. The combination of the direct photometric information [1] and the edge features leads to higher tracking accuracy and enables the approach to deal with challenging texture-less scenes. In contrast to traditional line feature based methods [2], we involve all edges rather than only line segments, avoiding the aperture problem and the uncertainty of endpoints. Instead of explicitly matching edge features, we design a dense representation of edges to align them, bridging the direct methods and the feature-based methods in tracking. Image alignment and feature matching are performed in a general framework including both pixels and salient visual landmarks. To track the camera pose, every new frame Fc is aligned to a reference frame Fr which is a carefully selected keyframe. First, the visual edges are extracted in Fc and Fr. Then, error caused by camera pose at Fc is estimated: nonedge points p in Fr are reprojected to Fc using