Master Thesis – Methods for Characterizing Optics and Image Reconstruction in LWIR Sensors

Thermal sensor pixels differ in sensitivity and offset, with further variations caused by ambient temperature and drift over time, creating noise that requires calibration and correction. This thesis at Axis Communications aims to model microbolometer pixel responses while accounting for optical effects, temperature and thermal drift. The findings could help simplify calibration and support future thermal cameras that operate without shutters.

The assignment

• Study the constraints of shutter-free thermal sensor operation.
• Investigate methods for correcting fixed-pattern noise.
• Assess how far sensor calibration can be performed without optics.
• Develop a model that predicts pixel responsivity across ambient temperatures, incorporating optical effects and drift over time.
• Build a camera prototype using existing Axis products to demonstrate the findings.

Who we are looking for

• Two master's students completing a thesis worth 30 credits each.
• Programming proficiency in at least one of Python, C++ or C is required.
• Courses in computer vision or image analysis are advantageous.
• Knowledge of optical science is beneficial.

What we offer

• Supervision from Leif Gellersen and Jimmy Heimdal.
• An opportunity to contribute to future thermal camera development through modelling and prototype work.
• A friendly, open and collaborative working culture.

Practical details

Location: Lund. Scope: two students, 30 credits per student, in fixed-term thesis positions. Application deadline: 27 November 2026. Questions: supervisors Leif Gellersen (leif.gellersen@axis.com) and Jimmy Heimdal (jimmy.heimdal@axis.com), or manager Daniel Larsson (daniel.larsson2@axis.com).