Essays about: "feature-based registration"
Showing result 1 - 5 of 8 essays containing the words feature-based registration.
-
1. Investigation of Registration Methods for High Resolution SAR-EO Imagery
University essay from Linköpings universitet/DatorseendeAbstract : With advancements in space technology, remote sensing applications, and computer vision, significant improvements in the data describing our planet are seen today. Researchers want to gather different kinds of data and perform data fusion techniques between them to increase our understanding of the world. READ MORE
-
2. Automatic Alignment Detection and Correction in Infrared and Visual Image Pairs
University essay from Linköpings universitet/DatorseendeAbstract : Previously well aligned image sensors, mounted on the same camera, might become misaligned depending on external vibrations. It is of interest to be able to automatically detect and correct for this misalignment, and to separate the deviation into pointing- and/or parallax errors. READ MORE
-
3. Implementation and Evaluation of a Variety of Image Stitching Methods
University essay from Uppsala universitet/Institutionen för informationsteknologiAbstract : Image stitching includes image registration and image merging. Image registration can be categorized into area based, frequency based, feature based, and—a recent addition—learning-based methods. As it is not straightforward to assess stitching accuracy, a metric is adopted for measuring stitching error. READ MORE
-
4. Automatic detection of manipulated packages by image comparison
University essay from KTH/Skolan för datavetenskap och kommunikation (CSC)Abstract : Shipping companies have a lot of difficulties to control and screen packages such as mail or different types of cargo that may contain dangerous or illegal goods. This project has the goal to enhance the supply chain security by elaborating a way to automatically detect manipulated packages. READ MORE
-
5. Überatlas: Robust speed-up of feature-based registration and multi-atlas based segmentation
University essay from Lunds universitet/Matematik LTHAbstract : Registration is a key component in multi-atlas approaches to medical image segmentation. Current state-of-the-art uses intensity-based registration methods, but such methods tend to be slow and sensitive to large amount of noise and anatomical abnormalities present in medical images. READ MORE
