Modelling dark current and hot pixels in imaging sensors
share
A gaussian mixture model was fitted to experimental data recorded under darkness by a camera for capturing astronomical images in order to model the distribution of hot pixels and dark current as functions of temperature and duration of exposure. The model accounts for random noise and lack of uniformity within the sensor; both components of variance are allowed to depend on experimental conditions. The results indicate that dark current does not grow linearly with respect to duration of exposure nor exponentially with respect to temperature, as claimed in the literature. In addition, interactions between the two experimental conditions are substantial. The results also provide interesting insights on the effect of temperature and duration on noise and lack of uniformity in the sensor.
READ FULL TEXT