自适应校正.docVIP

Scene-based Self Adaptive Iteration Bad Pixel Dynamic Correction Algorithm for Infrared Focal Plane Arrays Shuo Yang, Baojun Zhao, Linbo Tang School of Information and Electronics Beijing Institute of Technology Beijing, China yangdingshe@ Abstract—Infrared focal plane array IRFPA bad pixel correction in a variety of applications and complex conditions is a common problem. Therefore, a scene-based self Adaptive Iteration bad pixel dynamic correction algorithm was proposed and investigated. Adaptive filtering bad pixel detection algorithm is that the intensity of an image is recorded and the intensity of the image is filtered by a locally acting adaptive filter. Then the bad pixel regarded as an image noise is determined using intensity of the unfiltered image and intensity of the filtered image. Using the improved Nagao filtering bad pixel compensation algorithm with some stronger directivity filtering sub-windows whose patterns are decided by the texture analysis and which have the same pixel amount, the image boundary pixels did not reappear as flicking dot during the quick scene changes, and the rich scene details were preserved. Keywords-infrared focal plane array IRFPA ; scene-based method;bad pixel detection and compensation; texture analysis; nagao filters Introduction Infrared focal plane arrays IRFPA have been widely used in various fields. Due to the limitations of techniques and materials, the bad pixel always exists in kinds of IRFPA, which becomes an important problem effecting the image quality of IRFPA[1]. Usually, there are two steps of bad pixel correction: bad pixel detection and bad pixel compensation. According to the different generation mechanism of bad pixel, the detection methods are divided into static detection and dynamic detection. Static detection is suitable for the natural bad pixels of device, and it is based on calibration methods. A bad pixel map is set by calibration stage in laboratories or in factories, according to whic