红外图像中的脉冲噪声影响温度测量及分析,同时影响红外图像的可观性。传统中值滤波容易造成过度滤波,影响温度分布,丢失图像温度信息。针对传统方法的缺陷,借鉴噪声检测的中值滤波方法,将其应用于红外图像的脉冲噪声滤波。该方法在中值滤波之前进行一次脉冲噪声检测,确定受到噪声污染的像素点,并进行记录标识。然后进行对噪声污染像素的针对性滤波,实验证明,达到了既能有效滤除噪声,又完整保留温度信息的目的。

Impulse noise in infrared image may affect temperature measurement and analysis, and it also has effect on observability of the infrared image.Traditional method of median filtering can easily lead to excessive filtering, which can affect temperature distribution and cause lost of the temperature information.Considering the limitation of the traditional method, the median filtering method based on noise detection was used to deal with the impulse noise filtering of infrared image.Impulse noise checking was made before median filtering to confirm the pixels polluted by noise, which were recorded and marked.Then the pixels polluted by noise were filtered accordingly.Experiment proved that the method can filter out the noise effectively while keeping complete temperature information.