I'm trying to detect joints with color markers on a human for a robot.I know how to filter out the background using threshold.Now i want to find their position and detect them as objects.I'm new to opencv so please provide a little detail if possible.
Two possible solutions as far as i see, if you have indeed a binary image with black being background and white the dots or vice versa. 1) Blob detection, then for each detected blob visualize center - http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_feature_detectors.html?highlight=blob#SimpleBlobDetector%20:%20public%20FeatureDetector 2) Some type of region segmentation. No idea if this is actually available inside OpenCV. Know how to do this in Matlab, but go through documentation, it should be included.
Cheers
Asked: 2013-02-28 08:39:51 -0600
Seen: 840 times
Last updated: Feb 28 '13
Could you post an image sample of the markers?
thanks for the super fast reply man.cant do that.but it's fairly simple.it's all black with white spots which are fairly large with no visible noise.By the way I don't need to follow the objects seperately right now.I mean it's not important which is which.Just need the center coordinate of allof them in each frame.