Modeling event camera responses improves dynamic optical system analysis
Modelling dynamic systems transfer functions from events in computational neuromorphic imaging
Computer Vision and Pattern Recognition
Summary
Event vision sensors only respond to changes in light rather than steady illumination, making traditional testing methods invalid for these devices. The authors study how different types of light changes, like steps or ramps, affect the sensor’s output and how well these outputs can be used to understand the system’s behavior. They find that some light change patterns are better at overcoming sensor mismatches and noise than others. This work helps clarify how to model and interpret event camera data for better imaging system designs. It points toward building better tools that link scenes, sensors, and their dynamic responses.
What this means in practice
- •For imaging system engineers: Improve design and calibration of event-based cameras by choosing optimal temporal light probes for accurate system characterization despite sensor mismatches and noise.
- •For computer vision developers: Create algorithms that better interpret event camera data by understanding how dynamic optical signals affect measured outputs in neuromorphic imaging setups.
Authors
Nimrod Kruger, Gregory Cohen
Abstract
Event Vision Sensing (EVS) report threshold crossings of log-irradiance, so a static optical system imaging a static scene produces no output at all. The classical procedure for measuring a Point Spread Function (PSF), illuminating the system with a constant point source, therefore has no event-based equivalent: the probe must carry a temporal profile, and that profile becomes part of the measurement. A growing body of Computational Neuromorphic Imaging (CNI) work already exploits this, pairing engineered or modulated optics with event sensing, but each system adopts a particular excitation together with a particular reading of the event stream without the correspondence between the two being stated. We examine that correspondence directly within a analytical framework of an Linear Shift-Invariant (LSI) optical system with a specified Modulation Transfer Function (MTF), a first-order filter EVS pixel model, and three different temporal probes: a step function, a linear ramp and an exponential ramp. By analysing the inverse of the entire chain for different event-statistic, and comparing the results to the specified MTF, we identify the context where each probe is most relevant. We consider how photon-noise and cross-array threshold mismatch effects the analytical accuracy of the probe-inverse. Results show that the widely used step probe is highly susceptible to mismatch while resilient to photon shot-noise, while a linear rise probe and exponential rise probe retain their ability to infer signal levels even with high mismatch. We discuss the potential of dynamic-PSFs as components of a full forward operator from scene to events. In this, we use this analytical description to define dynamic-PSFs around EVS, and discuss the gaps toward a unified pixel model and a scene-composition framework required for CNI.