Fig. 1. Flow chart of the CICS-NUC method.

Fig. 2. Image data obtained by a scanning IRFPA with different ICS output（e.g. ICS equivalent output is 6000）：the original space image data superposed with the ICS radiance，FPN，and random noise.

Fig. 3. Images（f （x， y））obtained before the NUC by superposing various influence factors and a controllable steady-mode ICS equivalent output at 500，2000，3000，5500，and 6000.

Fig. 4. Figure 3′s outlier is marked during the filtering process.

Fig. 5. Evaluation images for the validation process: (a) gray-scale baseline equivalent output of 1500, NU = 19.15%; (b) gray-scale baseline equivalent output of 5000, NU = 15.87%.

Fig. 6. With the ICS radiation at 2000 and 6000 as the calibration points, the correction results of Fig. 5b are evaluated by different algorithms: (a) global mean two-point correction algorithm; (b) local mean two-point correction algorithm.

Fig. 7. Non-uniformity evaluation before and after NUC of each test image.

Fig. 8. NUC evaluation of test images by using different segmentation interval local mean algorithms.

Fig. 9. Implementing the proposed CICS-NUC algorithm on an actual space scene.

Fig. 10. NUC evaluation of test images by using MICS-NUC and CICS-NUC algorithms.