Dark background correction of the infrared detector for hyperspectral remote sensing application

The dark current is one of the main factors that influences the performance of infrared photodetectors, as it restricts the integration time and decreases the signal-to-noise ratio of an imaging system. Especially for hyperspectral remote sensing applications, the dark background occupies most of the dynamic range, owing to the narrow bandwidth and weak incident energy of each spectral channel. A general method for suppressing the dark current is to keep the optical system and the focal plane array (FPA) at an extremely low temperature (less than 30 K). However, maintaining a low temperature consumes a large amount of energy. Here, we present one correction method to eliminate the influence of the dark current on image quality. According to the real-time FPA temperature, the instantaneous dark background of each pixel is calculated and subtracted from the original image. Moreover, multiframe superposition is employed to improve the performance of the infrared hyperspectral imager. Based on this correction method, the noise equivalent temperature difference (NETD) in the imaging system is better than 0.3 K after 32 frames are merged.