Infrared Texture Simulation Using Gaussian-Markov Random Fields

Infrared battlefield simulation is a most important subject today to test infrared weapons. The reality and details of simulated scenes are always relied on the infrared textures. A texture is considered to be a stochastic, possibly periodic, two-dimensional image field. A texture model can be described as a mathematical procedure. Markov random field(MRF) model is a famous model to synthesize the visible textures of various kinds. In this paper, the Gaussian-Markov random fields(GMRF) are firstly used to sample the distribution of temperature field on the surface of terrain. And the Planck's radiation law is applied to calculate the emittance in the bands of 3～5 μ m or 8～14 μ m. The emittance distribution of the terrain can be normalized and colored by the range of 0 to the gray levels decreased by 1. Then an infrared texture is generated by GMRF by given a set of certain parameters.