本文为印度Rourkela国立技术研究所（作者：PRATISH KUMAR SAHOO）的硕士论文，共61页。

图像配准是将不同时间、不同视点和/或不同传感器拍摄的同一场景的图像（两个或多个）叠加的过程。图像配准将两个图像（参考图像和测试图像）在几何上对齐。本文讨论了利用互信息相关进行医学图像配准。提出了基于互信息的医学图像配准中特征匹配阶段存在的问题。匹配结果表明，通过改变图像的窗口大小（可以根据图像大小增减），可以得到更多的匹配点。本文还研究了两幅旋转失准图像的配准，给出了利用互信息测量旋转角度的方法。通常，配准应用于遥感、图像拼接、天气预报、超分辨率图像生成、医学（计算机断层扫描CT）和磁共振成像（MRI）的结合，以获得更完整的患者信息、肿瘤生长等。这里讨论了医学图像和卫星图像的拼接。论文最后总结了应用上述技术所取得的成果，并提出了今后为取得更好的效果可能做出的改进。

Image registration is the process ofoverlaying images (two or more) of the same scene taken at different times,from different viewpoints, and/or by different sensors. Image registrationgeometrically aligns two images (the reference and test images). This thesisdiscusses about the medical image registration by using correlation with mutualinformation. It also mentions about the problematic issue that arises duringthe feature matching stage in the mutual information based medical imageregistration. The matched results shows that by changing (can be decreasing orincreasing depending on the image) the windows size of the image more matchedpoints can be obtained. This thesis also focuses on registration of two imageswhich are rotationally misaligned and gives the procedure to measure this angleof rotation using mutual information. Typically, registration application arein remote sensing, image mosaicing, weather forecasting, superresolution imagecreation, in medicine (computerized tomography (CT) and magnetic resonanceimage (MRI) combining in order to obtain more complete information about thepatient, monitoring tumor growth etc. This thesis describes image mosaicing byusing mutual information. Here image mosaicing has been discussed for bothmedical and satellite images. The thesis concludes with results obtained byapplying above mentioned techniques and also provides with the possible changesthat could be made in future for obtaining better results.

1.  引言

2. 图像配准

3. 基于MI的医学图像配准

4. 旋转失准图像的配准

5. 参数随窗口大小的变化

6. 基于互信息的图像拼接

7. 结论

更多精彩文章请关注公众号：