A New Non-Uniformity Correction Method for Gradient Scenes Based on Scanning

The non-uniformity correction(NUC) algorithms are mainly divided into two categories, namely calibration-based NUC algorithms and scene-based NUC algorithms. They are generally based on some basic assumptions, generating significant limitations in practical engineering applications. Considering the airborne infrared imaging detection system can utilize its mechanism to scan the scene for detection, a scanning-based pure cloudless scene estimation method is proposed. The method consists of the following four steps: Control the motion of the mechanism of the infrared imaging detection system; Mark the motion pixels; Estimate the clean background estimation images; Obtain the gain and offset coefficients. A comparison of the uniformity of the images pre- and post-correction indicates the new method significantly improves the uniformity of the original images without additional manual intervention. It is easy to implement in the engineering context and meets the real-time requirements for practical engineering applications.