本文为芬兰米凯利应用科技大学（作者：Su Kai）的信息技术学士论文，共39页。
随着技术的进步，今天手机拍摄的照片比几年前拍摄的照片质量要好得多。这意味着，在许多年以后，我们可能并不满意今天所拍摄的手机照片。此外，摄影是一件好事，但绝不是一件简单的事情。有很多事情需要考虑，如ISO、白平衡、饱和度、标定等；大多数时候，我们并没有获取到一幅很好的图片。
本文的主题是从基本的角度分析这类需要增强质量的照片。直方图、FFT等工具都具有自己处理这些糟糕图像的特殊功能。从直方图中，我们可以看出亮度分布；通过FFT，我们知道图像的频域分布。作为JPEG文件的缺点，该类图片本身就不是很好，一些有用的信息已经丢失了。因此，手动调节这些像素值是一个很好的尝试方法。实际上，在本论文中，结合了一些数学知识来使用这些工具。例如，设计一个椭圆滤波器以适应图像的频谱图，以减少其中存在的瑕疵。
With technology improving, the photos taken by today’sphone are better than the phone photos taken years ago. It means that we maybenot satisfied with the photos taken by today’s camera phone years later.Another thing, shooting is a good but never be a simple thing. There is a lotof things to consider such as ISO, White balance, Saturation, Metering and soon. Most of time, we fail to get a good image.
The topic of this thesis is analysis this kind of photoin a basic way. Histogram, FFT. Each tool has its own function against thesebad images. From histogram diagram, we see the brightness distribution. By FFTwe know something about the frequency domain of a image. As the drawback ofJPEG file, the picture itself is not good, some useful information has lostsince the image is made. So manually managing these values of the pixels is agood way to try. Actually, in my thesis, I combine some mathematics when thesetools are underway. Such as, make a ellipse filter to fit the frequencyspectrogram to reduce more unexpected things.
下载英文原文地址：
http://page2.dfpan.com/fs/8lcj0221929106e8cb8/
更多精彩文章请关注微信号：