红外图像非均匀性校正和增强技术研究工学硕士学位论文.doc

工学硕士学位论文 红外图像非均匀性校正和增强技术研究 摘 要 由于红外成像系统具有抗干扰性能强、目标识别能力强、被动工作等特点，已被广泛的应用于军事和民用领域，是具有广阔发展前景的一种成像技术。因此，研制出高性能的红外成像系统具有重要的意义。 红外焦平面阵列是成像系统的最重要的器件，它的性能直接影响着整个系统的运行。但是，受到当前技术水平和工艺水平的限制，红外焦平面阵列存在严重的非均匀性，并且红外图像具有对比度低、目标与背景区分不明显的红外图像具有整体亮度偏、对比度较低、目标与背景区分不明显的特点 ABSTRACT Infrared imaging system has the characteristic of strong anti-jamming performance ，target recognition capability and also passive work,. It has been widely used in military and civilian fields. The imaging technique has broad prospects for development. So, there is important significance for the development of a high performance infrared imaging system. Infrared focal plane array is a key part of thermal imaging system. Its performance affects the operation of the system directly. However, limited by current technology, infrared focal plane arrays has the problem of non-uniformity, and infrared images have the features of low contrast; it is hard to distinguish between target and background. This paper presents many effective methods to solve these problems which exist in the infrared focal plane arrays. In this paper, more contexts are given: (1) The results obtained by traditional neural network correction algorithm, have the problems of blurring effect and some existing artifacts. After analysis the reasons of the problems that the traditional neural network correction algorithm may cause, in this article an improved algorithm is proposed: replace the mean filter, which used in the traditional algorithm by the nonlinear filter. And the parameters of nonlinear filter are variable according to the local image detail changes in the improved algorithm. The corrected images by the improved algorithm achieve not only a significant improvement in image clarity, but also effectively eliminate the problem of artifacts caused by the traditional algorithms. (2) After analyzed the cause of blind pixels and the difference between blind pixels and point targets, an effective detection algorithm was proposed which