Non-uniformity Correction For Infrared Images Without A Shutter

With the gradually development of infrared thermal imaging technology,it has been widely used in military,civil and other fields.However,because of the limitation of material and manufacturing process,there usually remain some differences between infrared detectors in a focal plane array(FPA),which would cause obvious non-uniformity noise in infrared images.Besides the non-uniformity noise is more affected by the image scene,and is also influenced by the FPA temperature in uncooled infrared imaging system,resulting in the severe decrease of image quality.So achieving non-uniformity correction(NUC)and nullifying the effect of external factors to image quality stability is an important link of infrared imaging system.In the traditional NUC methods,the correction parameters is usually steady,without consideration of the image scene and FPA temperature.So the imaging quality of infrared system is not always satisfactory.To solve this problem,a shutter is typically used as an external uniform temperature source to calibrate and update correction parameters regularly.However,the shutter not only require higher cost and power,but also interrupt the imaging process temporarily.Therefore,some NUC methods base on shutter-less technology have been proposed in recent years.But most of these algorithms tend to simply consider the effect of the PFA temperature,so they cannot get excellent corrected results.Grounding two primary factors impacting the non-uniformity,we present a shutter-less method for the NUC of infrared images on the theoretical basis of existing NUC methods.This method consists of two phases: NUC based on the focal plane array temperature and NUC based on the image scene information.The first step is to estimate correction parameters by a temperature-dependent interpolation based on the current FPA temperature,and then achieve a real-time gain and offset correction.The second step is to calculate adaptively correction parameters based on the image scene,and then conduct a further compensation correction.The proposed algorithm is evaluated and compared with other representative NUC algorithms objectively and subjectively,using real infrared images containing different scenes.Experimental results show that this method can not only reduce the non-uniformity noise effectively,but also update correction parameters adaptively according to the change of the FPA temperature and the image scene,therefore a certain stability of the imagequality can be obtained.