Modelling dark current and hot pixels in imaging sensors. (February 2020)

• Record Type: Journal Article
• Authors: Forcina, Antonio; Carbone, Paolo
• Is Part Of: Statistical modelling. Volume 20:Number 1(2020)
• Page Start: 30
• Page End: 41
• Publication Date: 2020-02
• Subjects: dark current -- hot pixels -- dark frames -- Gaussian mixtures -- components of variance -- latent class models; Linear models (Statistics) -- Periodicals; Mathematical models -- Periodicals; Modèles linéaires (Statistique) -- Périodiques; Modèles mathématiques -- Périodiques; Modèle statistique; Modèle linéaire; Modélisation statistique; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 519.5011
• DOI: 10.1177/1471082X18803464
• Languages: English
• ISSNs: 1471-082X

Abstract

A Gaussian mixture model with a structured covariance matrix was used to analyse image data recorded by a digital sensor under darkness to model the effects of temperature and duration of exposure on the expected value and on the variance of the sensor dark current, separately for ordinary and possibly defective pixels. The model accounts for two components of variance within each latent type: random noise in each image and lack of uniformity within the sensor; both components are allowed to depend on experimental conditions. The results seem to indicate that the dependence of the expected value of dark current on duration of exposure and temperature cannot be represented by a simple parametric model. The latent class model detects the presence of at least two types of hot pixels. If we order the latent classes in decreasing order of the class weights, the corresponding expected values and variances increase. The covariance structure that emerges from our analysis has an important implication: the sign and the relative size of pixels deviations from uniformity are invariant to experimental conditions.