Thank you Sir for the information.

Dear krishna,

You can utilise the Kalman Filter for object tracking which is one of the best techniques for  estimation and tracking. Some inputs  can be helpful such as :

1) https://in.mathworks.com/help/vision/examples/using-kalman-filter-for-object-tracking.html

The above link provides a MATLAB implementation you can try.

The following paper might be useful :

http://file.scirp.org/pdf/IIM_2013032920521334.pdf

Hi,

I think you mentioned two problems: obeject tracking and distance calculation.

Object tracking: the best way way to track one object is to find the special features on it. If you can mark the object, sphere is a very common mark used in object tracking. If you can, you can try some feature extraction algorithms, like SIFT. Anyway, if you can find the features of this object, you can find the matched one in each frame.

Distance calculation: first of all, a single camera is hard to find the distance of the object. We usually use the double-camera system or the infrared-camera system to calculate the distance of one object, like a depth scanner. 

Yu Luo

Thanks Shubham. I will go through it and get back to the post

Thanks Yu Luo for the information. I didnot get you regarding the distance calculation. Actually I used high speed camera to capture the video. So it contains around 10000 frames.

Hi,

After you tracking the object in your frames, you will get the path of the object. But it's only the path of pixels, it's not the physical distance. If you want to convert these pixels data into physical data, like calculating the velocity of the object, you will need a stereo-vision system.

Yu Luo