Kanji Tanaka

About

Kanji Tanaka is a robotics researcher whose work centers on mobile robot localization, mapping, and autonomous navigation in dynamic, real-world environments. His most significant contribution lies in developing robust, scalable algorithms that enable robots to accurately determine their position even in large, changing environments — a notoriously difficult challenge in autonomous systems research. Tanaka's most cited work (136 citations) pioneered the application of support vector machines to RFID-based robot self-localization, demonstrating how inexpensive identification technologies could be harnessed for reliable positioning. He has also made substantial contributions to RANSAC-based relocation techniques, introducing both incremental and LSH-accelerated variants that allow robots to operate efficiently at scale. His LSH-RANSAC and Monte Carlo localization frameworks address the computational bottlenecks posed by high-dimensional feature matching — a critical concern as robotic maps grow increasingly complex. Beyond localization, Tanaka has explored dynamic environment mapping, vision-based obstacle avoidance by observing pedestrian movement, and multi-robot systems. His research consistently bridges theoretical algorithmic innovation with practical deployment constraints. With over 200 cumulative citations across his body of work, Tanaka has established himself as a meaningful contributor to the foundational literature of autonomous mobile robotics.

Research Focus

Key Achievements

Top Papers

1. 1
   Self-localization of mobile robots with RFID system by using support vector machine

   136 citations · 2005
2. 2
   LSH-RANSAC: An incremental scheme for scalable localization

   29 citations · 2009
3. 3
   Incremental RANSAC for online relocation in large dynamic environments

   28 citations · 2006
4. 4
   Detecting changes in a dynamic environment for updating its maps by using a mobile robot

   11 citations · 2002
5. 5
   Mobile robot localization with an incomplete map in non-stationary environments

   10 citations · 2004
6. 6
   Detecting collision-free paths by observing walking people

   9 citations · 2003
7. 7
   A scalable algorithm for monte carlo localization using an incremental E<sup>2</sup>LSH-database of high dimensional features

   8 citations · 2008
8. 8
   On the scalability of robot localization using high-dimensional features

   7 citations · 2008
9. 9
   A Scalable Localization Algorithm for High Dimensional Features and Multi Robot Systems

   6 citations · 2008
10. 10
    A vision system for detecting mobile robots in office environments

    6 citations · 2004

Key Collaborators

Contact & Links