Hi Mr. Alireza,

               I have mentioned few references for your kind look at it. I believe the materials present in these documents will help you a lot in getting a detailed answer of your concern. Feel free to ask any more doubts.

Regards,

A.Kar

http://www.academia.edu/1542615/Hand_Detection_and_Tracking_Using_Depth_and_Color_Information

http://link.springer.com/chapter/10.1007%2F978-3-642-38577-3_59

http://link.springer.com/chapter/10.1007%2F978-3-642-55038-6_76

http://www.cs.ucsb.edu/~mturk/pubs/KolschFG2004.pdf

http://www.admin.cam.ac.uk/offices/research/documents/local/events/downloads/ict/Roberto_Cipolla.pdf

http://ethesis.nitrkl.ac.in/2887/1/209ec1111.pdf

kinect is the sensor which may be helpful

hi, you can check this out: https://www.youtube.com/watch?v=m4qv9rK8k9s

the utilized approach is based on blob detection

Your question is about two quite different topics. Detection of an object in an image and subsequent tracking of the object. There is a great deal of ilterature on both of these topics, e.g. tens of thousands of papers in the IEEE. The detection part is more or less template matching at a basic level but there could be a probabilistic part to it since a hand is not an object of fixed size and shape. The tracking part will depend on whether you wish to track just the centroid of the hand or individual parts of the hand that have a more or less constant geometrical relationship to each other. I suggest you start with IEEE Xplore and do keyword searches, sorting by date from earliest to latest, to start with. Some classic books could also help - Duda and Hart, Bar-Shalom and Fortmann, etc.

Recently, Convolutional Neural Networks are used heavily for this field.

Example from Google:

https://ai.googleblog.com/2019/08/on-device-real-time-hand-tracking-with.html

OpenPose (Not so sure if tracking is involved):

https://github.com/CMU-Perceptual-Computing-Lab/openpose