Scene-based nonuniformity correction based on bilateral filter with reduced ghosting

• We determine whether the movement of the scene exists to decide to update the parameters or not.
• We decrease the update rate of correction factors in high brightness regions.
• We speed the update of correction factors in regions with brightness decreased sharply.

In order to correct the nonuniformity noise of the IR imaging system, a nonuniformity correction deghosting algorithm based on high-brightness region detection was proposed. Firstly, the bilateral filter was used to filter the image. The high-brightness regions in the result image were detected, and the correction factors corresponding to the regions were reduced. And then, the positions were detected which are high-brightness pixels in former frame but are not in current frame, and the correction factors corresponding to the regions were enlarged. Thus, the wrong update of the correction factor caused by high-brightness region of both the current frame and the previous frame was reduced. The correction method was validated using a real image sequence. The experimental results showed that the suppression effect of the proposed algorithm on the ‘ghost’ effect caused by the high-brightness region is better than the other two algorithms named SLTH-NUC and BFTH-NUC, respectively. Moreover, the proposed algorithm can also suppress the ‘ghost’ effect caused by the sharp edge of the scene.