A Microbolometer-Based Time-to-First-Spike Pixel for Infrared Imaging

This paper presents the first implementation of a Time-to-First-Spike (TFS) pixel based on an uncooled microbolometer for infrared (IR) imaging. To the best of our knowledge, no TFS pixels based on thermal sensors have been demonstrated, while TFS pixels have been explored in the visible spectrum using photodiodes. In this work, the microbolometer is modeled as a temperature-dependent resistor whose voltage is compared against a reference ramp. The crossing time, which varies with temperature, is captured as a digital spike using a C-element and converted to a digital signal via a Time-to-Digital Converter (TDC). This temporally encoded scheme enables efficient, event-driven readout and naturally suppresses spatial redundancy, as pixels responding to similar IR radiation spike simultaneously. To validate this approach, a proof-of-concept (PoC) testchip was designed in 130nm CMOS, integrating a single TFS pixel, a 3×3 TFS pixel array, a ramp voltage generator, and a bias current source. The readout circuitry, including the TDC and control logic, will be implemented on an FPGA to support experimentation with both global and rolling shutter strategies. This work opens a new path toward low-power, event-based IR sensors and neuromorphic thermal imaging systems.