Cade Wells


Smart vision in SoC applications

STMicroelectronics (2000-04)

Research Engineer: Cade Wells
Sponsor: STMicroelectronics
Supervision: Dr David Renshaw and Dr Iain Lindsay,
University of Edinburgh

STMicroelectronics holds a world leading position in image sensors with primary markets in webcams and digital still cameras, and is also a market leader in the production of image processors for set-top boxes. The company believes that a growing proportion of future mainstream imaging applications will require embedded ‘smart’ image processing that could track or recognise objects in real time. Example applications for this type of smart camera could include airbag deployment or security via face recognition.

The research project will lay the groundwork for future realisation of this capability. Initially, a familiarity of up-to-date image processing will be acquired; including the macro functions, such as object tracking, and the low-level algorithms from which they are composed, such as convolvers. A core set of operators in a range of styles, delivering different cost/time trade offs will then be identified and implemented. ST expects to use and build upon this preliminary library for real products.

The Research Engineer will then advise on, and possibly automate the choice issue, generating automatically the style required to meet the constraints of a given implementation. Some generic system architectures may emerge which are capable of addressing many applications from the same hardware base.

  • Research

  • Login




  • Click to visit our educational website

  • Archives