Difference between revisions of "ICP software"
| Line 1: | Line 1: | ||
Iterative Closest Point (ICP) is a widely used method to match two sets of points related by a rigid-body transformation. For example, you are imaging a sample using confocal microscopy before and after some treatment, and you want to be able to realign your sample in order to compare the two image stacks. | Iterative Closest Point (ICP) is a widely used method to match two sets of points related by a rigid-body transformation. For example, you are imaging a sample using confocal microscopy before and after some treatment, and you want to be able to realign your sample in order to compare the two image stacks. | ||
| − | In this [http://www.springerlink.com/content/x44h731l56642l72/ paper], we provided a novel method to perform ICP by using an iterative estimation scheme. The source code for this method is available here: [[Media:Icp.tar.gz.]]. This code assumes relatively good quality data, and does not handle partially overlapping data sets. | + | In this [http://www.springerlink.com/content/x44h731l56642l72/ paper], we provided a novel method to perform ICP by using an iterative estimation scheme. The source code for this method is available here: [[Media:Icp.tar.gz.]]. This code assumes relatively good quality data, and does not handle partially overlapping data sets. It was developed on linux but should work on windows and mac as well. It uses no fancy libraries. |
Revision as of 09:33, 20 April 2011
Iterative Closest Point (ICP) is a widely used method to match two sets of points related by a rigid-body transformation. For example, you are imaging a sample using confocal microscopy before and after some treatment, and you want to be able to realign your sample in order to compare the two image stacks.
In this paper, we provided a novel method to perform ICP by using an iterative estimation scheme. The source code for this method is available here: Media:Icp.tar.gz.. This code assumes relatively good quality data, and does not handle partially overlapping data sets. It was developed on linux but should work on windows and mac as well. It uses no fancy libraries.
