Electroluminescence Radiance Maps based on Multiple Exposure Images from InGaAs Cameras

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

70 Downloads (Pure)


This paper presents a method to increase the dynamic range of the EL signal for diagnostics of defective areas in PV modules. It applies a computational method for reconstructing a high dynamic range radiance map from a stack of standard dynamic range images at multiple exposure times designed by Debevec and Malik for visible light cameras (RGB). Increasing the precision of the EL signal has two interests: more quantitative information is then available for automatic analysis, and better visual quality is available for visual inspection in case of a small amplitude EL signal, such as acquired in an outdoor daylight setup using lock-in EL. Both are demonstrated via two experiments during which a PV module is electrically biased and imaged at 8 different exposure times with an InGaAs camera: one indoor and one outdoor in daylight. The exposure times are combined to reconstruct an EL image of higher precision. To measure the increased precision, the number of unique pixel values is used: it increases by a factor larger than 2.5 for the indoor case and a factor higher than 100 in the outdoor case. In the outdoor case the improvement towards visual inspection is also clearly apparent.
Original languageEnglish
Title of host publicationProceedings of the 8th World Conference on Photovoltaic Energy Conversion
PublisherEU PVSEC
Publication date2022
Pages448 - 453
ISBN (Print)3-936338-86-8
Publication statusPublished - 2022
Event8th World Conference on Photovoltaic Energy Conversion - Milano Convention Centre, Milano, Italy
Duration: 26 Sept 202230 Sept 2022
Conference number: 8


Conference8th World Conference on Photovoltaic Energy Conversion
LocationMilano Convention Centre
Internet address


  • Electroluminescence
  • Defects
  • Experimental methods
  • Evaluation
  • Reliability


Dive into the research topics of 'Electroluminescence Radiance Maps based on Multiple Exposure Images from InGaAs Cameras'. Together they form a unique fingerprint.

Cite this