A multi-fractal formalism for stabilization, object detection and tracking in FLIR sequences

Shekarforoush, H.; Chellappa, Rama

doi:10.1109/icip.2000.899299

Cited by 30 publications

(23 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Situations of this type arise, for example, in military applications where the object signatures and dynamics may be poorly known or even totally unknown prior to the initiation of tracking, where the sensor platform may be subject to strong ego-motion relative to the imaged objects and backgrounds, and where it is infeasible to compensate for the ego-motion in real-time via registration techniques such as those given in [13,16] due to insufficient computational bandwidth and/or insufficient inertial measurements. Template tracking is commonly applied directly to the video frames acquired from an imaging sensor, where the peak of the normalized correlation between the template and local neighborhoods of the most recently acquired frame is This work was supported in part by the U.S. Army Research Laboratory and the U.S. Army Research Office under grant W911NF-04-1-0221.…”

Section: Introductionmentioning

confidence: 99%

Modulation Domain Template Tracking

Nguyen

Havlicek

Yeary

2007

2007 IEEE Conference on Computer Vision and Pattern Recognition

View full text Add to dashboard Cite

For the first time, we perform normalized correlation template tracking in the modulation domain. For each frame of the video sequence, we compute a multi-component AM-FM image model that characterizes the local texture structure of objects and backgrounds. Tracking is carried out by formulating a modulation domain correlation function in the derived feature space. Using visible and longwave infrared sequences as illustrative examples, we study the performance of this new approach relative to two basic pixel domain correlation template trackers. We also present preliminary results from a new dual domain tracker that operates simultaneously in both the pixel and modulation domains.

show abstract

Section: Introductionmentioning

confidence: 99%

Modulation Domain Template Tracking

Nguyen

Havlicek

Yeary

2007

2007 IEEE Conference on Computer Vision and Pattern Recognition

View full text Add to dashboard Cite

show abstract

“…Unfortunately, there are some disadvantages for the infrared image, such as extremely low signal to noise ratio (SNR), severe background clutter, non-repeatability of the target signature, and high ego-motion of the sensor. Commonly, many infrared object tracking algorithms are imposed different constraints, such as no drastic change of the object feature [7], and no sensor ego-motion [8]. In this paper, we follow the 'mean shift' tracking approach, which was proposed by D. Comaniciu et al [5,6].…”

Section: Introductionmentioning

confidence: 99%

“…Infrared object tracking primarily addresses to localize and track the infrared thermal object in the infrared image sequences [7,8]. Unfortunately, there are some disadvantages for the infrared image, such as extremely low signal to noise ratio (SNR), severe background clutter, non-repeatability of the target signature, and high ego-motion of the sensor.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Infrared Object Tracking Based on Mean Shift

Cheng

Yang

2004

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The mean shift algorithm is an efficient method for tracking object in the color image sequence. However, in the infrared object-tracking scenario, there is a singular feature space, i.e. the grey space, for representing the infrared object. Due to the lack of the information for the object representation, the object tracking based on the mean shift algorithm may be lost in the infrared sequence. To overcome this disadvantage, we propose a new scheme that is to construct a cascade grey space. The experimental results performed on two different infrared image sequences show our new scheme is efficient and robust for the infrared object tracking.

show abstract

“…The K-mean method is then used to classify the target and background based on the resultant fractal dimensions. Shekarforoush et al, (17) have also used a multi-fractal formalism for object detection and tracking in FLIR sequences.…”

Section: Introductionmentioning

confidence: 99%

Adaptive target detection in FLIR imagery using the eigenspace separation transform and principal component analysis

et al. 2003

View full text Add to dashboard Cite

Sensors and Electron Devices Directorate, ARLApproved for public release; distribution unlimited.ii REPORT DOCUMENTATION PAGE Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. REPORT DATE (DD-MM-YYYY)June 2005 ARL-TR-3544 SPONSOR/MONITOR'S ACRONYM(S) 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) U.S. Army Research Laboratory 2800 Powder Mill Road Adelphi, MD 20783-1197 SPONSOR/MONITOR'S REPORT NUMBER(S) DISTRIBUTION/AVAILABILITY STATEMENTApproved for public release; distribution unlimited. SUPPLEMENTARY NOTES ABSTRACTIn this paper, an adaptive target detection algorithm for forward-looking infrared (FLIR) imagery is proposed which is based on measuring differences between structural information within a target and its surrounding background. At each pixel in the image a dual window is opened where the inner window (inner image vector) represents a possible target signature and the outer window (consisting of a number of outer image vectors) represents the surrounding scene. These image vectors are then preprocessed by two directional highpass filters to obtain the corresponding image gradient vectors. The target detection problem is formulated as a statistical hypotheses testing problem by mapping these image gradient vectors into two linear transformations, P 1 and P 2 and, via principal component analysis and eigenspace separation transform, respectively. The first transformation P 1 is only a function of the inner image gradient vector. The second transformation P 2 is a function of both the inner and outer image gradient vectors. For the hypothesis H 1 (target): the difference of the two functions is small. For the hypothesis H 0 (clutter): the difference of the two functions is large. Results of testing the proposed target detection algorithm on two large FLIR image databases are presented. SUBJECT TERMStarget detection, eigenvector analysis, eigenspace separation transform, principal component analysis, FLIR imagery, statistical hypotheses testing SECURITY

show abstract

A multi-fractal formalism for stabilization, object detection and tracking in FLIR sequences

Abstract: ABSTRACT

Cited by 30 publications

References 11 publications

Modulation Domain Template Tracking

Modulation Domain Template Tracking

Real-Time Infrared Object Tracking Based on Mean Shift

Adaptive target detection in FLIR imagery using the eigenspace separation transform and principal component analysis

Contact Info

Product

Resources

About