Scene-Based Adaptive Iteration Bad Pixel Dynamic Correction Algorithm for Infrared Focal Plane Arrays

Abstract

Infrared focal plane array (IRFPA) bad pixel correction as a prevalent problem is widely used in complex conditions. Therefore, scene-based adaptive iteration bad pixel dynamic correction algorithm is proposed and explored. Adaptive filtering bad pixel detection algorithm is to record the intensity of an image and replace the intensity of the image filtered by a locally acting adaptive filter. Using the intensity of the unfiltered and filtered image to determined the bad pixel which regarded as image noise. By the improved Nagao filtering bad pixel compensation algorithm, filtering sub-windows with the same pixel amount whose patterns are decided by the texture analysis can obtain stronger directivity. Thus, the image boundary pixels do not reappear as flicking dot during the quick scene changes, and the rich scene details are preserved.