【作者】 朱建；

【导师】 朱东波； 刘贵喜；

【作者基本信息】 西安电子科技大学 ， 机械制造及其自动化， 2007， 硕士

【摘要】 红外热成像技术作为一种发现、探测和识别目标的重要手段无论是在军事上,还是商业中都有着广泛的应用。红外焦平面成像系统则是今后红外成像技术发展的趋势。但红外焦平面各阵列元响应的非均匀性严重限制了红外成像系统的成像质量,因此对红外焦平面阵列进行图像处理,获取高质量的红外图像成为红外成像技术应用的关键之一。本文对红外图像处理算法进行研究。首先介绍了焦平面阵列非均匀性产生的原因及国内外目前的几种非均匀性算法:全局非均匀性校正、时域高通滤波、加权均值滤波算法、神经网络算法、两点线性校正和多点分段线性校正等算法。其中重点对两点线性校正和多点分段线性校正算法进行了研究,并在这两种算法的基础上提出了一种新的算法:基于两点分段线性校正的改进算法。通过对前几种算法的仿真相比较,取得了更好的校正效果。在对图像进行了非均匀性校正后,还要进行图像增强处理,本文系统研究了常用图像增强算法。根据红外图像特性,提出了增强高通滤波的算法,同时将增强高通滤波和平台直方图均化两种图像增强的结果进行加权平均。并对算法进行了仿真,该算法图像增强实时性强,图像增强效果好。更多还原

【Abstract】 Infrared imaging technology is widely used in both military field and civilian field by means of finding, detecting and identifying the targets. Imaging system of infrared focal plane arrays (IRFPA) developed into the main trend in future. But the uniformity occurred by IRFPA is badly constrained the image quality of the infrared imaging system. So one of the hinges in the application of infrared image technology is to get higher quality of the infrared image.Some infrared image processing algorithm is studied in the paper. Firstly, the reason that led to the nonuniformity of the focal plane array and several current nonuniformity algorithms, such as full-scale nonuniformity correction; high-pass filtering; weight average filtering, nerval networks filtering, two-point linearity correction and multi-point multi-section linearity correction are studied home and abroad. And two-point nonuniformity correction algorithm and multi-point multi-section linearity correction algorithm are paid emphasis on. A new algorithm, the improved algorithm based on the two-point multi-section linearity correction algorithm, is presented based on the comparison of two algorithms above. Comparing the several algorithms above, better correction result is gotten.更多还原