本文为美国托莱多大学（作者：MengLi）的硕士论文，共88页。
近年来，数字图像处理在各个领域得到了广泛的研究和应用。作为数字信号处理的一个应用方向，数字图像处理比模拟图像处理具有许多优点；它允许对输入数据使用更广泛的算法，并且可以避免在处理过程中出现诸如噪声和信号失真等问题。在本文中，我们将介绍三种处理数字图像的重要算法：图像去噪、图像增强和目标检测与跟踪。本文提出的遗传算法（GA）可以检测和跟踪弱小目标、低可观测目标和点目标，主要用于远程监控应用。
作为更有效地检测和跟踪对象的第一步，首先对输入图像进行降噪和增强。我们采用全变差（TotalVariation，TV）技术去除噪声，提高输入图像的信噪比（SNR）。为了进一步增强用于户外应用的图像，引入了雾天图像增强技术，该技术显著地有益于交通和户外视觉系统。雾天图像增强是数字图像处理的一个重要分支，它是在有雾的情况下使用的。为了克服现有雾天图像增强算法的不足，本文提出了一种结合主成分分析（PCA）、多尺度同态滤波（MSR）和全局直方图均衡（GHE）的图像增强方法。首先，将PCA变换应用于雾状图像，将输入图像分割成亮度和两个色度分量。在第二步中，分别通过MSR和GHE增强亮度和色度分量。最后，应用逆PCA变换将三个通道的结果组合成新的RGB图像。
在数字图像序列中检测并跟踪低可观测目标，设计了一种遗传操作的编码方案来跟踪目标。为了避免丢失任何目标轨迹，引入个体保存方法来维持更有希望的候选轨迹，然后通过多阶段假设检验方案来确认目标轨迹。
Over the past few years, digital imageprocessing has been widely studied and used in various fields. Image processinguses computer algorithms to perform image processing on digital images. As asubcategory or field of digital signal processing, digital image processing hasmany advantages over analog image processing. It allows a much wider range ofalgorithms to be applied to the input data and can avoid problems such as thebulid-up of noise and signal distortion during processing. In this thesis, weare going to introduce three important algorithms dealing with digital images:image denoising, image enhancement and target detection and tracking. Theproposed Genetic Algorithm (GA) can detect and track dim, low observable andpoint targets, mainly for remote monitoring applications. As a first step todetect and track objects more effectively, the input image is first denoisedand enhanced. We use Total Variation (TV) technique to remove the noise andimprove the Signal to Noise Ratio (SNR) of the input image. To further enhancethe image for outdoor applications a foggy image enhancement technique isintroduced which significantly benefits traffic and outdoor visual systems.Foggy image enhancement is an important branch of digital image processing,which is used when the weather is foggy. To overcome the shortcomings of theexisting foggy image enhancement algorithms, we have developed a method that combinesPrincipal Component Analysis (PCA), Multi-Scale Retinex (MSR) and Global HistogramEqualization (GHE). Initially, a PCA transform is applied to the foggy image tosplit the input image into a luminance and two chrominance components. In the secondstep, the luminance and the chrominance components are individually enhanced byMSR and GHE, respectively. In the final stage, an inverse PCA is applied tocombine the results of the three channels into a new RGB image. To detect andtrack low observable targets in a digital image sequence, an encoding schemealong with genetic operation is designed to track the targets. To avoid missingany tracks, individual preservation method is introduced to maintain the more promisingcandidate tracks. Target trajectories are then confirmed by a multi-stage hypothesistesting scheme.
1 引言
2 历史文献回顾
3 基于全变差的图像降噪技术
4 基于主成分分析的图像增强技术
5 基于遗传算法的低可观测目标检测与跟踪
6 仿真结果
7 结论与未来研究工作
附录A 全变差图像降噪的源代码
附录B PCA图像变换的源代码
附录C 基于MSR的亮度分量增强源代码
附录D 基于GHE的色度分量增量源代码
附录E 目标检测源代码
附录F 目标跟踪源代码