红外焦平面阵列盲元检测及补偿算法

Algorithm of Blind Pixels Auto-searching and Compensation for IRFPA

为了消除非制冷红外焦平面阵列图像盲元、改善成像质量,利用红外焦平面阵列盲元相应特性和相邻像元的相关性,提出了一种基于阈值+相邻像元检测及加窗中值补偿算法。结果表明,阈值+相邻元检测算法实现盲元误检率低、快速查找等优点;利用加窗中值的盲元补偿算法,可以使盲元图像得到明显改善,保持图像信息的完整性,同时,算法易于硬件实现。

In order to eliminate blind pixel of the infrared focal plan arrays and improve the image quality, this paper put forward one category of arithmetic about the threshold value and the adjacent pixels detection and adding window median repair by using IRFPA＇s image data of two reference radiation source and the pertinence between the adjacent pixels. The result unfolds the exeellences about the threshold value and the arithmetic of adjacent pixels detection, such as the lower misdiagnosis rate of blind pixel, the faster lookup etc. It can get obvious improve of the blind pixel picture and retain the image＇s information by using the arithmetic of adding window median repair, and, this arithmetic is prone to actualize on hardware.