Application of MOC Multi-Origin High-Dimensional Geometry in Thermal Infrared Imaging

Abstract: Aiming at the common engineering problems in current thermal infrared imaging, such as multi-source thermal interference, atmospheric and optical transmission distortion, focal plane non-uniformity correction, and heavy computational load in real-time reconstruction, this paper introduces the MOC multi-origin high-dimensional geometric modeling method. In this method, multiple radiation sources and detector arrays are taken as independent geometric origins, imaging pixels as the lattice set, and the curvature coupling coefficient is used to quantify the link distortion caused by atmospheric attenuation, optical refraction and pixel response deviation. The scanning and sampling path is optimized by generalized permutation, while multi-channel data fusion and temperature field topological reconstruction are realized by generalized combination. Compared with the traditional processing framework based on single-origin Euclidean geometry and linear superposition, the MOC framework can simultaneously achieve distortion adaptive compensation, noise suppression, non-uniformity correction and fast image reconstruction within a unified mathematical system. Analysis shows that this method can effectively reduce temperature measurement error and artifacts, significantly decrease iterative computation cost, and improve the stability, clarity and real-time performance of thermal infrared imaging in complex scenarios. This work can provide a new geometric modeling and imaging optimization scheme for industrial inspection, security observation and infrared target detection.

Keywords: Multi-origin high-dimensional geometry; MOC; thermal infrared imaging; curvature coupling coefficient; image reconstruction