Infrared multi-detector non-uniformity images simulation method

With the rapid development of aerospace technology, there is an increasing demand for wide-field, high-resolution infrared detector systems. The focal plane size has evolved from centimeters to sub-meters, and the size of a single detector cannot meet the requirements. Consequently, the inevitable trend is the mosaic integration of multiple detectors. Research on image processing associated with multi-detector mosaicking necessitates obtaining raw images from as many detectors as possible. However, due to cost constrains, it is crucial to simulate a large number of images based on a limited amount of detector image data. This paper improves upon existing methods for simulating the original images of single detectors. By establishing correlations between the model coefficients of pixel response functions, the simulation of single detector images closely aligns with real data compared to existing methods. Simultaneously, the arrangement characteristics of readout circuit channels are considered during parameter generation, resulting in simulated images that visually resemble real images more closely. Furthermore, this paper proposes a method for simulating original images of an arbitrary number of detectors based on the analysis of real data from a limited number of detectors of the same batch. The method generates multi-detector images that closely match the magnitudes, trends and ranges of real data in terms of mean, std (standard deviation), and non-uniformity. Additionally, the visual perception of these images closely resembles that of real multi-detector scenarios.