Emerging infrared and visible imaging applications require higher sensitivity, larger pixel array, larger contrast ratio (dynamic range), very low power consumption and faster data readout rate operations all at the same time. Some of these applications are camera surveillance used both in day/night (very bright and dark conditions), medical diagnostics, weather forecasting, and aerial search & rescue operations etc. The digital-pixel focal plane array (DFPA) implemented in this thesis has the capabilities to capture a wide dynamic range of more than 120dB in a single global shutter without saturating the pixels at a huge frame rate of more than 500Hz. An adaptive Integration Window technique has been developed which ensures that we are able to measure such a huge dynamic range using a counter of only 10 bits (this helps us lower the power consumption of the design). This proposed image sensor has been designed, fabricated and tested in 65nm CMOS technology. It has 16 x 16-pixel array with 16 x 9 pixels with an inbuilt Silicon APD for optical testing and 16 x 7 dummy pixels for electrical testing. Our design proposes an off-chip digital calibration technique to cut down the burden on the analog circuitry. The sensor design achieved more than 128dB+ of dynamic range with a DNL/INL of 0.65/1.65 respectively with a power consumption of only 0.58 uW/pixel. The digital calibration scheme successfully cuts down the pixel-pixel variation standard deviations by a factor of 4. The proposed image sensor design should be able to address most of the short-comings of conventional FPAs and provides a one-shot solution to the design of high performance CMOS image sensors.
