Harvey E. Rhody, Ph.D.

Ph.D., Electrical Engineering
Syracuse University

3144 Carlson
E-Mail: rhody@cis.rit.edu

Research Highlights

Signal & Image Processing

The extraction and use of information from natural signals presents many important and interesting problems. Natural signals, such as speech and images, are rooted in nature. Other signals, such as radar, sonar, and ultrasound are created by the interaction of synthetic signals with natural processes and media. In both cases, the signals contain information that has been encoded by nature and which we wish to extract for some purpose. Natural adaptation has given us perceptual systems that are well matched to some natural signals such as sounds and images. The development of algorithms and systems to artificially extract useful information from such signals has proven to be exceedingly difficult, but it is a necessity for the development of advanced automation.

Recent developments in artificial neural networks, fuzzy logic, object-oriented programming, and progress in understanding human perception have opened new approaches to intelligent signal processing. Development of more powerful computers and new computer architectures provides a new level of technology. In recent years we have made progress in speech understanding, the application of artificial neural networks to image segmentation and vowel recognition, the use of fuzzy logic for information fusion, and the use of object-oriented programming to speed software development. These techniques can now be combined to develop algorithms for many practical applications.

Recent projects in our intelligent signal processing laboratory have included the development of artificial neural network algorithms for image segmentation, a speech research workstation, a multi-media information system using a networked object-oriented database, and object-oriented discrete event simulation system. We are currently working on a visual programming tool for signal and image processing, an expert system utilizing fuzzy logic for the control of a factory energy management system, an image segmentation system based on a cascade-correlation neural network. A computer vision workstation under development will provide the foundation for work in image understanding. I would like to meet students who are interested in developing algorithms using advanced computer techniques and artificial intelligence for natural signal processing applications.

Go to SIMG-782 Digital Image Processing class page