This is a multiparticle version of the CMorph algorithm previously submitted. The algorithm takes the gradient of the image, and then applies decreasing thresholds and fill operations to attempt to identify objects. One it has found the objects, it attempts to pair them up with objects in the previous frame. It was designed to track cells and microbeads in flow cell assays, so it is designed to be biased to look for a flow direction.
The program will output two files:
1. A color-coded *.avi with object numbers so it is easy to check the results.
2. A *.mat file with the tracking results stored as a structure. The field id's will match the object numbers in the video.
I hope this tool will help you in your research. If you use this to help with a publication, please cite the work in your methods section. Either my dissertation or the CMorph conference publication, both are listed in the readme file. I will update the references with the appropriate journal paper as soon as it is in press.