Pixel-Level Dual-Gain Infrared Image Reconstruction Based on Radiance Calibration

Abstract

Nonuniformity is a key indicator of the image data quality of a remote infrared sensor. The uniformity correction residual of the image obtained by a pixel-level, dual-gain, time delayed and integration(TDI) detector is closely related to the linearity after reconstruction. In this study, the signal readout process of the infrared TDI detector was analyzed. A new image data reconstruction method is presented to obtain accurate normal parameters for each pixel. The detector output linearity was increased in all dynamic ranges, and the uniformity of the image was enhanced. Radiance calibration was performed, and the test data were processed. The result shows that the nonuniformity correction residual decreased from 4.1% to 1.2% based on radiance calibration.