Robust autonomous detection of the defective pixels in detectors using a probabilistic technique

Area of Science:

• Sensor technology
• Image processing
• Astronomy instrumentation

Background:

• Defective pixel detection in solid-state detectors is an under-researched area.
• Existing methods for detecting defective pixels in real images are limited and ad hoc.
• There is a need for a robust and widely applicable technique for defective pixel detection.

Purpose of the Study:

• To present a novel probabilistic integrated technique for autonomous detection of defective pixels in image sensor arrays.
• To develop a method applicable to various digital cameras and detector types.
• To address the limitations of current defective pixel detection approaches.

Main Methods:

• A probabilistic integrated technique is proposed for autonomous defective pixel detection.
• The method is designed to be applicable to images with rich scene information from any digital camera.
• The technique can identify different types of defective pixels within the detector.

Main Results:

• The technique was successfully applied to detect various defective pixels in an experimental camera with a charge-coupled device (CCD) array.
• The method was also validated on HgCdTe detectors used in the UKIRT's Wide Field Camera (WFCAM) for infrared astronomy.
• Demonstrated effectiveness in identifying defective pixels in real-world astronomical imaging scenarios.

Conclusions:

• The developed probabilistic technique offers an effective solution for autonomous defective pixel detection.
• The method shows versatility, applicable to different detector technologies (CCD, HgCdTe) and imaging applications.
• This approach advances the field by providing a robust and broadly applicable tool for sensor quality assessment.