Bayesian Signal Processing and Algorithm Development

We are leaders in the application and implementation of Bayesian analysis to systems where conventional "text book" solutions do not exist, conventional signal processing assumptions do not hold or where optimal operation is a necessity.

This approach enables us to optimise the partitioning between hardware and software within complex systems, leading to improved performance or simplification of the overall product.

The application of this development approach has led to a number of innovative products and improvements in existing systems:

• World first in particle sizing sensor for industrial applications
• Compact and efficient audio watermarking system with robust performance
• High accuracy and low-cost missile test and development system
• Chemical and drug detection systems with improved resolution
• Inspection system with significantly improved false alarm rate

Our Bayesian analysis experience has been used to support complex sensor fusion work where data from multiple sensor systems has to be combined and analysed. This further strengthens the range of our skills in conventional image and radar processing. It also has resulted in low power, cost-effective sensor fusion solutions for the automotive market.