Larry Matthies

About

Larry Matthies is a pioneering researcher in robotic perception, computer vision, and autonomous navigation, whose work has fundamentally shaped how robots see and move through the world. Best known for his foundational contributions to stereo vision and visual odometry, Matthies established rigorous probabilistic frameworks for depth estimation and landmark tracking that remain cornerstones of the field. His early papers on Kalman filter-based depth estimation (712 citations) and 3D error modeling in stereo navigation (501 citations) introduced statistically principled approaches that dramatically improved mobile robot positioning accuracy. Matthies's most celebrated achievement is his central role in deploying Visual Odometry on NASA's Mars Exploration Rovers — the first time such capability was demonstrated on another planet — enabling rovers to autonomously detect and compensate for wheel slip across alien terrain (642 citations). His broader body of work on planetary robotics, including stereo vision systems and terrain learning for slip prediction, helped define the discipline of space robotics. Beyond planetary exploration, he has contributed to sensor fusion, moving object detection from autonomous vehicles, and first-person activity recognition for human-robot interaction. With thousands of citations spanning nearly four decades, Matthies stands as one of the most influential figures in field robotics and autonomous perception.

Research Focus

Key Achievements

Top Papers

1. 1
   Kalman filter-based algorithms for estimating depth from image sequences

   712 citations · 1989
2. 2
   Two years of Visual Odometry on the Mars Exploration Rovers

   642 citations · 2007
3. 3
   Error modeling in stereo navigation

   501 citations · 1987
4. 4
   First-Person Activity Recognition: What Are They Doing to Me?

   294 citations · 2013
5. 5
   Stereo vision for planetary rovers: Stochastic modeling to near real-time implementation

   257 citations · 1992
6. 6
   Rover navigation using stereo ego-motion

   246 citations · 2003
7. 7
   Computer Vision on Mars

   185 citations · 2007
8. 8
   Integration of sonar and stereo range data using a grid-based representation

   175 citations · 2003
9. 9
   Learning and prediction of slip from visual information

   150 citations · 2007
10. 10
    Real-time detection of moving objects from moving vehicles using dense stereo and optical flow

    134 citations · 2005

Key Collaborators

Contact & Links