File:Automated-Cell-Identification-and-Tracking-Using-Nanoparticle-Moving-Light-Displays-pone.0040835.s005.ogv

English: Pseudo-noise filter. Investigation of the effects of increased noise is possible by varying the cut-off of the noise filter during the image cleaning process. The graphic demonstrates the effects of this variation on a single time frame. The blue dots correspond to binary elements, the red dots correspond to the centroids identified through clustering of the binary elements and the green dots mark the manually identified locations of the cells. As the filter is lowered the binary element clusters increase to sufficient numbers for each cell to allow accurate centroid allocations. As the filter approaches the noise floor of the image an increasing number of noise peaks are wrongly identified as binary element markers and these randomly occurring features result in the miss-allocation of centroids.