CACS FACULTY AND THEIR RESEARCH INTERESTS

Dr. CHEE-HUNG HENRY CHU

hchu @ cacs.louisiana.edu

Professor of Computer Engineering.
B.S.E. summa cum laude, Computer Engineering, Univ. Michigan, Ann Arbor, 1981;
M.S.E., Computer, Information and Control Engineering, Univ. Michigan, Ann Arbor, 1982;
Ph.D., Electrical Engineering, Purdue Univ., West Lafayette, Indiana, 1988.

Click here for contact information.

Dr. Chu's research interests are in pattern recognition, image processing, and machine vision. One recent project draws from both pattern recognition and image processing to develop steganalytic methods that detect hidden messages in image data. A detecting method can be shown to be more powerful than the embedding method if the detector uses a more accurate model of the image or video. Previous work either have been based on using simple classification rules on features extracted from images, or have used powerful classification methods such as the support vector machine on simplistic image processing results, such as computing the wavelet transform coefficients. The ongoing collaboration between Dr. Chu and Dr. Michael Pratt demonstrates the impact of the amount of texture on the detectability of hidden messages and proposes a solution that improves the detection performance in low embedding rate applications.

Another pattern recognition project is in combining different data sources in medical informatics for prediction the onset of diseases. A recently started project is a collaboration with GE Healthcare to study the prediction of Alzheimer's disease using imaging data collected using different modalities, viz. positron emission tomography and magnetic resonance imaging, together with non-image-based data, such as the patient's demography, lifestyle, and clinical assessments.

The second major component of his work is in using image and video data to understand the three-dimensional (3D) world. His interest in machine vision is in creating 3D models of existing structures, which currently is a labor-intensive and time-consuming task. 3D content generation has applications in visualization, immersive environments, and video game development. The key to achieving automatic and efficient content generation is via digital photogrammetry, which is the process of performing non-contact measurement through digital image analysis. The extracted measures are used to generate 3D models that match the real-world counterparts.

PROFESSIONAL ACTIVITIES

Dr. Chu is a Registered Professional Engineer in Electrical Engineering. He is a member of the Institute of Electrical and Electronics Engineers and the Association for Computing Machinery, and the honor societies of Tau Beta Pi, Eta Kappa Nu, and Phi Kappa Phi. In May 2000, he received the university foundation's Distinguished Professor Award. In 2003, he was recognized by the Louisiana Board of Regents for his service to the Speaking of Science program. In 2006 and 2008, he was recognized by the university's Phi Kappa Phi honor society chapter for his service.

In the summer of 1997, Dr. Chu was a visiting professor at the computer engineering department of Kasetsart University, Bangkok, Thailand. He was a visiting professor at the computer science department of the Prince of Songkhla University, Hatyai, Thailand, in the summer of 2005.

SELECTED PUBLICATIONS

S. Chiewchanwattana, C. Lursinsap, and C.H. Chu, "Imputing incomplete time-series data based on varied-window similarity measure of data sequences," Pattern Recognition Letters, vol. 28, no. 9, pp. 1091-1103, 2007.
K. Sunat, C. Lursinsap, and C.H. Chu, "The p-recursive piecewise polynomial sigmoid generators and first-order algorithms for multilayer tanh-like neurons," Neural Computing and Applications, vol. 16, no. 1, pp. 33--47, Springer, Jan. 2007.
Y. Zhang and C.H. Chu, "Inverse polar-ray projections for recovering projective transformations," in Proc. IEEE International Conf. Computer Vision and Pattern Recognition, Anchorage, Alaska, June 2008.
L. Wang and C.H. Chu, "Graph-theoretic segmentation of airborne lidar data," in Proc. Conf. Independent Component Analyses, Wavelets, Unsupervised Nano-Biomimetic Sensors and Neural Networks VI, SPIE Symposium on Defense and Security, Orlando, Fla., March 2008.
Q. He, R.R. Schultz, C.H. Chu, "Efficient superresolution image reconstruction applied to surveillance video captured by small unmanned aricraft systems," in Proc. Conf. Signal Processing, Sensor Fusion, and Target Recognition XVII, SPIE Symposium on Defense and Security, Orlando, Fla., March 2008.
Q. He and C.H. Chu, "Intrinsic images from Fisher linear discriminant," in Proc. International Symposium on Visual Computing, Lake Tahoe, Nev., Nov. 2007.
Y. Zhang and C.H. Chu, "One-dimensional mappings for recovering large scale projective transformations," in Proc. IEEE Int. Conf. Acoustics, Speech, and Signal Processing, Honolulu, Hawaii, Apr. 2007.