Jiangbo Lu
Papers
2
Total Citations
90
H-Index
2
About
Dr. Jiangbo Lu is a leading figure in real-time computer vision, with a primary focus on high-performance stereo matching and parallel computing. His most influential work centers on the development of algorithms that achieve both exceptional accuracy and real-time processing speeds, a critical challenge for applications like autonomous navigation and 3D reconstruction. Dr. Lu’s major contribution is a scalable, GPU-accelerated approach for stereo matching, as detailed in his seminal 2009 and 2011 papers, which have together garnered over 90 citations. He pioneered a GPU-oriented bitwise fast voting method that dramatically improves matching accuracy without sacrificing speed. By combining this with cost aggregation over shape-adaptive support regions, his work demonstrated that local algorithms could rival the accuracy of slower global methods while running in real-time on CUDA-enabled hardware. This breakthrough has made high-quality depth estimation practical for embedded and real-world systems. Dr. Lu’s research is notable for bridging the gap between theoretical accuracy and practical, real-time performance, establishing a foundational methodology that continues to influence modern stereo vision systems.
Research Focus
Key Achievements
Top Papers
- 1Real-time accurate stereo with bitwise fast voting on CUDA46 citations · 2009
- 2