File Exchange

image thumbnail

Lucas-Kanade Tracker with pyramid and iteration

version (6.35 KB) by Ke Yan
Use Lucas-Kanade Algorithm to track feature points between 2 images


Updated 27 Jul 2012

View License

The file contains Lucas-Kanade Tracker with pyramid and iteration to improve performance. There is a wrapper for image sequences, and a corner detection function using Shi-Tomasi method.

Comments and Ratings (12)

siva ganesh


I guess you should use original X1,Y1 instead of computed xt,yt for the calculation of Ix Iy I1 in each higher pyramid.



Can some one please help me a demo or steps to execute and how to use test data to observe the results.

Theo Walker

Nice! This works well for my cell tracking application.

If you want to learn more about the method, take a look at the lecture that this code is based on:


Hi Edward,
could you tell me what literature you used for this algorithm?
Was it by any chance:
1. Pyramidal Implementation of the Lucas Kanade Feature Tracker
Description of the algorithm?

Best Regards


Does not works well for moving camera.


Hi friend i have to track cars in the vedio (camera is also in motion).... Kindly tell how to start with your algorithm and whether it works


Hi Edward,

How do you use your algorithm?
Is there any demo?

I'm able to see what to do with X2 and Y2?

Best Regards

Ke Yan

Hi Edward,

Is there any test data for us to play?

Kindly Regards,


.zip file

fixed a bug reported by Bronislav Přibyl, function eig2 in corner_ST.m, detA = a(1,1,:).*a(2,2,:)-a(1,2,:).*a(2,2,:);

Add an Email

MATLAB Release Compatibility
Created with R2010b
Compatible with any release
Platform Compatibility
Windows macOS Linux