single record

Algorithm/Service Info

Algorithm or Service name
Cluster Crawler
Author or Maintainer
Bruce Baller
one line description
Reconstructs line-like 2D clusters, 2D vertices and 3D vertices.

The Cluster Crawler algorithm constructs clusters using local information instead of a global metric for pattern recognition such as what is done in the Hough Line algorithm. The crawler is similar to the technique of gliders in 2D cellular automata. Short line-like “seed” clusters of proximate hits are constructed in an area of low hit density where hit proximity is a good indication that the hits are indeed associated with each other. Additional nearby hits are attached to the cluster end if they are similar to the hits already attached to the cluster. Seed clusters are formed at one end of the hit collection so that crawling in only one direction is sufficient.

location in code
2D hit reconstruction, other
code analysis done
improved code released

Author Info