DSOL: A Fast Direct Sparse Odometry Scheme

In this paper, we describe Direct Sparse Odometry Lite (DSOL), an improved version of Direct Sparse Odometry (DSO) [1]. We propose several algorithmic and implementation enhancements which speed up computation by a significant factor (on average 5x) even on resource-constrained platforms. The increase in speed allows us to process images at higher frame rates, which in turn provides better results on rapid motions. Our open-source implementation is available at https://github.com/versatran01/dso1.

Authors

• Chao Qu
• Shreyas S. Shivakumar
• Ian D. Miller
• Camillo J. Taylor

Keywords

• Multicore processing
• Intelligent robots
• Direct Sparse Odometry
• Frame Rate
• Parallelization
• Linear System
• Directed Graph
• Depth Images
• High Gradient
• Inertial Measurement Unit
• Current Image
• Pixel Coordinates
• Map Points
• Direct Alignment
• Stereo Images
• Sum Of Squared Differences
• Frames Per Second
• Reprojection Error
• Bundle Adjustment
• Visual Odometry
• Pyramid Level
• Factor Graph
• Stereo Matching
• Pose Error
• Local Map
• Entire Window
• Computational Resources
• Image Pyramid