Research On Infrared Image Expansion Methods Based On Measured Data

The infrared scene simulation method based on real measurement data is to use the real measurement data to measure the radiation and obtain the real radiation distribution of the target scene,and then apply the radiation measurement results to the infrared scene simulation to generate high precision infrared simulation images.Therefore,it is of great research significance to study various infrared image generation methods based on infrared radiation characteristics.In the acquisition of infrared images,there is a need to study infrared image expansion methods as the diversity of real infrared images is limited.IR image expansion techniques not only solve the problem of time-consuming and laborintensive real-world IR images but also provide the most realistic reproduction of the IR radiation properties of the target surface and generate highly realistic IR images.Therefore,in this paper,we analyze the infrared imaging characteristics of the target,determine the variation rules of the target radiation concerning distance,orientation,and time,and study the method of expanding the infrared image with high realism under different imaging distances,orientations and moments,and generate infrared images with different distances,orientations and moments to obtain high The method is used to generate infrared images with different distances,orientations and moments,and to obtain highly realistic infrared texture images.Firstly,the transmission characteristics of infrared radiation are analyzed from the principle of infrared imaging;the energy distribution characteristics model and the distance imaging relationship model are established from the target infrared imaging characteristics,the theory of infrared image orientation expansion is introduced,and the time variation law of infrared radiation is analyzed to provide the theoretical basis for the subsequent infrared image expansion.Secondly,the whole process of infrared radiation transmission and final infrared image generation is analyzed,as well as the various effects involved in the imaging system.The inversion method is used to remove atmospheric effects and various imaging system effects,and the single-band and double-band inversion methods are adopted to invert the target surface temperature respectively,and the accuracy of the inversion results is analyzed and optimized.Finally,the method of generating infrared image expansion for different distances,orientations,and times is investigated based on the real images.Based on the inversion results of the measured data,the method of IR image distance expansion is investigated by the inversion-reconstruction method in combination with the above-mentioned target IR imaging relationship characteristics and radiation energy distribution characteristics;the method of orientation expansion is investigated and the results are adjusted by using optical flow calculation to estimate the angle change of the scene according to the affine transformation;meanwhile,further work is done to generate IR images of different time Further extension work is done to generate highly realistic infrared images at different distances,azimuths and moments.