A novel optical pixel sensor (design on the left) has been developed in the dept. ETRO to measure the fluorescence life-time decay by emitting laser pulses onto the object (setup on the middle). The optical response can be used to detect e.g. changes in tissue characteristics in (bio)medical applications (measurement setup on the right).
The aim is to determine the decay characteristics (shown on the left) as accurately as possible as it characterizes the device. This characterization can be done accurately using system identification methods. However, to apply the method in practice in a camera, the numerical complexity needs to be reduced. The minimum is the 2-sample method (shown on the right).
The goal of this thesis is to study the spectrum of possible methods and implements, from the full-blown identification methods down to the 2 sample method, with “all” possibilities in between. After prototyping the methods at a higher level (e.g. in MATLAB), the “optimal”/selected method can be implemented on a FPGA for real-time processing.
Hans Ingelberts (ETRO)