Driver Image Recognition and Position Using Field Programmable Gate Arrays 2008-01-0787
The implementation of biometrics in the automotive industry still remains in its infancy. The factors hindering its maturity are breach of safety, cost, informal environment and variable ambient parameters. The use of biometrics is currently arguable as to its feasibility within an automobile, however its implementations are undoubtedly inevitable. Vehicle security, authorisation of ownership and human factor diagnosis are instances that would benefit from biometrics.
The proposal is to design and implement a system that will recognise the side-profile facial features, concentrating on the auricle of the ear of an individual. This method of choice was governed by the informal environment and safety aspect required. The design will cater for the filtration required to reduce the effects of variation in ambient lighting and other visual interferences such as a passenger, window reflection or passing traffic.
The method proposed is to capture a side profile image and then orientate to a best-fit correlation with a masked and filtered authorised image previously stored in memory.
The hardware used to carry out this task will predominantly consist of a Field Programmable Gate Array (FPGA) to carry out the captured image filtration and orientation. The results of a correlation with a stored image can then be sent onto the vehicle networks.