The Study On UWB SAR Change Detection Of Foliage-concealed Targets Based On Visual Attention

Due to Uitra Wide Band Synthetic Aperture Radar(UWB SAR) System working in low frequency, it has longer wavelength compared with high- frequency SAR system. At the same time, because of its strong ability in detecting foliage-target, the UWB SAR also plays an important role in civilian and military aspects. However with the resolution of the low-frequency UWB SAR system improved continuously, the low speed of processing become the biggest problem that we should to solve. Inspired by the visual attention mechanism and change detection's advantages, in this paper we combine the visual attention mechanism with change detection technology in the detection of UWB SAR Foliage-concealed targets.There are three aspects in my paper.First of all, we should know the basic theoretial of visiual attention mechanism. In this part, we review the physiological structure of human visual system, clarify the basic concepts, summarize the main characteristics, discuss t he feasibillitty of visual attention mechanism using in UWB SAR image. In order to provide necessary support for later research, we set up framework for visual attention model which is suitable for UWB SAR image.Secondly, we do some research about UWB SAR visualization. In this chapter, combined with the human visual characteristics, we analysis the statistical properties and the gray- level distribution model of UWB SAR image data. Futhermore, we come up with a new fast visualization method so that we can compress the UWB SAR data dynamic range reasonably.Lastly, we drop up a new method to detect the UWB SAR foliage-concealed target which is based on bottom- up visual attention mechanism. In this algorithm, multi- scale features are combined into a single visual saliency map by using the model of visual attention. Then the image local neighborhood information and spatial correlation are used to improve the performance of detection. Moreover, thel result shows that the algorithm can detect the foliage-concealed targets between multi-temporal UWB SAR images with less time and it is robust to registration error.