Single Infrared Image Non-uniformity Correction Based on Genetic Algorithm

Non-uniformity correction for single infrared image is typical multi-extremum problem, which can be solved by genetic algorithms suitably and efficiently. Based on the idea of genetic algorithms (GAs), in this paper a new single image stripe non-uniformity correction method has been established successfully. Non-uniformity correction parameters (gain and offset coefficients) are turned into individual of GAs. Special constraint are designed to form the fitness function, which is efficient for optimization process of GAs. To validate our method, comparisons with other traditional methods are presented through real infrared images. Experimental results demonstrate the proposed method can eliminate stripe non-uniformity of infrared images as well as high consistency of original image.