Tsutomu Maruyama
Papers
4
Total Citations
87
H-Index
4
About
Tsutomu Maruyama is a leading figure in real-time computer vision, renowned for his pioneering work in hardware-accelerated stereo matching. His research focuses on developing high-speed, accurate depth-sensing systems for mobile and embedded platforms, including autonomous robots, drones, and self-driving cars. Maruyama’s foundational contribution, "A Real-Time Stereo Vision System with FPGA" (2003), has garnered 55 citations and established a critical pathway for using reconfigurable hardware to achieve real-time performance. To address the power and resource constraints of mobile devices, he later pioneered algorithmic innovations like the Z2-ZNCC (ZigZag Scanning based Zero-means Normalized Cross Correlation) method. This approach, detailed in papers from 2020 and 2021, dramatically accelerates stereo matching on embedded GPUs, achieving both high accuracy and low latency—a breakthrough cited over 28 times. His recent work on video stabilization (2022) further extends his impact, enabling smoother footage from handheld and drone-mounted cameras in real-time. Through a career dedicated to bridging algorithmic efficiency with hardware constraints, Maruyama has enabled practical, high-performance vision systems that are essential for the next generation of autonomous and mobile technologies.
Research Focus
Key Achievements
Top Papers
- 1A Real-Time Stereo Vision System with FPGA55 citations · 2003
- 2Efficient stereo matching on embedded GPUs with zero-means cross correlation23 citations · 2021
- 3
- 4Acceleration of video stabilization using embedded GPU4 citations · 2022