The advent of Internet and the increased demand for multimedia (text, images, video and sound) by general public, business and industry have created a need for compression of material transmitted over the work-wide networks and for storage. Compression allows the same material to be transmitted faster over the Net, or more material can be stored in the same space. This laboratory has been involved in research useful to compression for a number of years, and is now leading in several areas.
We have developed compression algorithms capable of reducing the size of images and sound (both speech and music) considerably at a high quality. The techniques involve wavelets, self-similarity, genetic algorithms, neural networks, as well as fractal and multifractal characterization of signals.
We have developed applications such as identification of transients in radio transmitters and faults in power systems of "fingerprinting" of such transmitters and faults. We are working on identification of cancerous changes in breast cells. The algorithms are also applied to sequencing and classification of genetic material. We have developed a system for seeding depth monitoring and control for farms in Manitoba.
In collaboration with the Telecommunications Research Laboratories (TR Labs), we are developing techniques for analysis and control of broadband traffic on coaxial and fiber optics networks as well as low-bit-rate video.
* Large network of computers
* Sound acquisition
* Image acquisition
* Microscopic image acquisition
* Transient acquisition
Recent clients include:
* Department of Communication (Canada)
* Industry Canada
* Department of Education and Training (Manitoba)
* Environment Canada
* Institute of Electrical and Electronics Engineers
* Manitoba Hydro
* Linnet Geomatics
* Profile Management
For more information, contact:
Dr. W. Kinsner
+1 (204) 474-6490