Hey all!

I am currently working on a project to detect motion from people. For this a webcam is hanging above a surface. Simply, it detects motion by drawing black pixels where it sees motion. The blob detection recognised these black pixels and creates blobs. I then use the coordination of a blob (Xblob,Yblob) to see where the people are on the field.

However, people are often divided into multiple blobs (for an arm, head or feet for example). I want to create one blob from all these seperate blobs: a cluster of blobs = one blob if < maximum blob size of a human (variable). In other words:

1. I need to tell the maximum size of a blob that is, the maximum size of one person so that seperate blobs are used to create one blob.

For example: If a lot of blobs are detected near each other it is most obvious that these cluster of blobs are to large for just one person. Looking at the maximumblobsize for one person this should be two persons: creating two blobs.

I am having problems how to write code to create a single blob out of multiple blobs (I expect some kind of algoritm).


Any help/tips/feedback on this topic would be very much appreciated.
Kind regards,
Patrick