Determination of feature generation methods for PTZ camera object tracking

Daniel D. Doyle; Jonathan T. Black

doi:10.1117/12.918452

15 May 2012 Determination of feature generation methods for PTZ camera object tracking

Daniel D. Doyle, Jonathan T. Black

Proceedings Volume 8395, Acquisition, Tracking, Pointing, and Laser Systems Technologies XXVI; 83950O (2012) https://doi.org/10.1117/12.918452
Event: SPIE Defense, Security, and Sensing, 2012, Baltimore, Maryland, United States

Abstract

Object detection and tracking using computer vision (CV) techniques have been widely applied to sensor fusion applications. Many papers continue to be written that speed up performance and increase learning of artificially intelligent systems through improved algorithms, workload distribution, and information fusion. Military application of real-time tracking systems is becoming more and more complex with an ever increasing need of fusion and CV techniques to actively track and control dynamic systems. Examples include the use of metrology systems for tracking and measuring micro air vehicles (MAVs) and autonomous navigation systems for controlling MAVs. This paper seeks to contribute to the determination of select tracking algorithms that best track a moving object using a pan/tilt/zoom (PTZ) camera applicable to both of the examples presented. The select feature generation algorithms compared in this paper are the trained Scale-Invariant Feature Transform (SIFT) and Speeded Up Robust Features (SURF), the Mixture of Gaussians (MoG) background subtraction method, the Lucas- Kanade optical flow method (2000) and the Farneback optical flow method (2003). The matching algorithm used in this paper for the trained feature generation algorithms is the Fast Library for Approximate Nearest Neighbors (FLANN). The BSD licensed OpenCV library is used extensively to demonstrate the viability of each algorithm and its performance. Initial testing is performed on a sequence of images using a stationary camera. Further testing is performed on a sequence of images such that the PTZ camera is moving in order to capture the moving object. Comparisons are made based upon accuracy, speed and memory.

Citation Download Citation

Daniel D. Doyle and Jonathan T. Black "Determination of feature generation methods for PTZ camera object tracking", Proc. SPIE 8395, Acquisition, Tracking, Pointing, and Laser Systems Technologies XXVI, 83950O (15 May 2012); https://doi.org/10.1117/12.918452

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available