Study On Complex Background Suppression Algorithm For Infrared Surveillance Warning System

Infrared Surveillance Warning System (ISWS) is electro-optical information equipment which detect, recognise and track targets by receiving their infrared emission passively. The ISWS which has invisibility, high resolving ability and fine anti-amming ability. For these advantages, ISWS beomes an important component of the modern information warfare system, and obtains wide attention and energetic cooperation in recent years by militarys. Therefore, how to make the best of infrared target detection ability to increase the distance of target detection and obtain the related information, have become important significance to improve the performance of information warfare system. However, because of target along distance, there are small size, only serveral pixels and no shape, texture information in imaging plane for infrared camera and targets are submerged in complex background. Therefore, comparing with other topics in the field of infrared target detection and tracking, how complex backgrounds can be robustly suppressed under low signal-to-clutter ratio have becomes frontier research topics which have important theory significance and engineering application value.In this paper, based on comprehensive analysis for in-the-art infrared image background suppression technique and considering different emission intensity and distributed structure in infrared image between dim small target and background, complex background suppression techniques are used research deeper for key pre-processing technique of dim and small target detection and tracking. Multi-scale geometry analysis, local filter, statistics and variational partial differential equation theorys are analyzed and used, respectively, and then a seris of new infrared image dim and small target background suppresion algorithms are provided:(1)According to nonsubsampled contourlet transform (NSCT) characteristics with multi-scale, multi-directional and translation invariance. Develop two new infrared image dim and small target background suppression algorithms based on singularity value decomposition and fuzzy logic which adjusted the NSCT subbands coefficients with truncated eigenvalue and fuzzy nonlinear background suppression operator, respectively. The experimental results demonstrate that new algorithms can suppress the background of clouds fluctuating effectively, save and enhance target signal.(2)Analyze relationship each pixel with local neighbourhood pixel in spatial distances, intensity and local regions. Design and implement two new infrared image dim and small target background suppression algorithms based on multiresolution bilateral filter and multi-scale non-local means filter. The experimental results demonstrate that new algorithms can suppress the background of terrain fluctuating and ground road network effectively, save and enhance target intensity information.(3)Combing Bayesian maximum a posteriori estimation with shearlet transform. Design and implement a new dim and small target background suppression algorithm based on the Gaussian scale mixture model creatively. The experimental results demonstrate that the new algorithm can suppress complex background of artificial building for surface-sky detection system effectively.(4)Analyze the common algorithms of variational partial differential equation. According to characteristic with energy functional and multi-scale analysis. Construct and implement dim and small target background suppression algorithms based on RX operator improved anisotropic nonlinear diffusion equation and total variation Gabor model for artificial building of surface-sky detection system background suppression and keep the target signal steadily. Several experiments testify theirs practicability and validity. And then, total variation Gabor model algorithm is applied in ISWS with staring detector successfully.