Synthetic Aperture Radar Synthetic Aperture Radar Automatic Target Recognition Automatic Target Recognition -Computer Science Department- -Computer Science Department- California Polytechnic State University, California Polytechnic State University, San Luis Obispo San Luis Obispo Alvin Y. Wang Alvin Y. Wang and and Chia-Huei Yao Chia-Huei Yao Faculty Advisor: Faculty Advisor: Dr. John Saghri Dr. John Saghri Project Sponsor: Project Sponsor: Raytheon Company Raytheon Company Contact Personnel: Contact Personnel: Jeff Hoffner Jeff Hoffner
Synthetic Aperture Radar Automatic Target Recognition. -Computer Science Department- California Polytechnic State University, San Luis Obispo Alvin Y. Wang and Chia-Huei Yao Faculty Advisor: Dr. John Saghri Project Sponsor: Raytheon Company Contact Personnel: Jeff Hoffner. Agenda. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
IntroductionIntroduction Usage of image identificationUsage of image identification
MilitaryMilitary MedicalMedical
SAR imagesSAR images MSTAR image MSTAR image
databasedatabase
Courtesy of Sandia National Laboratory
Synthetic Aperture Radar Synthetic Aperture Radar SAR instruments use pulses of microwaves SAR instruments use pulses of microwaves
as an active source of illuminationas an active source of illumination BenefitsBenefits
Independent of light sourcesIndependent of light sources Capable to see through cloudsCapable to see through clouds Spatial resolution remains the same no matter Spatial resolution remains the same no matter
how far the target area ishow far the target area is
The light intensity variations on an image aThe light intensity variations on an image are caused by an object’s surface orientatiore caused by an object’s surface orientation, its reflectance, and characteristics of its n, its reflectance, and characteristics of its lighting sourcelighting source
Based on the variance of light intensity, we Based on the variance of light intensity, we can classify and group the underlying imagcan classify and group the underlying image into some topographical categoriese into some topographical categories
Based on the location of the topographical Based on the location of the topographical features, we can reasonably reconstruct thfeatures, we can reasonably reconstruct the original 3D model.e original 3D model.
Feature Extraction and Distance TransformFeature Extraction and Distance Transform
Image Matching procedureImage Matching procedure Find contour points of the reference shape and oFind contour points of the reference shape and o
btain their DTbtain their DT Obtain contour points of the measured shapeObtain contour points of the measured shape Compute and superimpose the centroids of the tCompute and superimpose the centroids of the t
wo point setswo point sets Rotate and translate the measured point set with Rotate and translate the measured point set with
respect to the initial poserespect to the initial pose Select those relative positions that yield the miniSelect those relative positions that yield the mini
mum HD valuemum HD value Select the one with the least mean HD.Select the one with the least mean HD.
CDT Provides good approximation to the exact CDT Provides good approximation to the exact Euclidean distanceEuclidean distance
Distance Trasform converts a binary image to another Distance Trasform converts a binary image to another image in which pixel value is the distance from this image in which pixel value is the distance from this pixel to the nearest nonzero pixel of the binary image.pixel to the nearest nonzero pixel of the binary image.
courtesy of IPAN
Image Matching procedureImage Matching procedure
Image Matching procedureImage Matching procedure Find contour points of the reference shape and oFind contour points of the reference shape and o
btain their DTbtain their DT Obtain contour points of the measured shapeObtain contour points of the measured shape Compute and superimpose the centroids of the tCompute and superimpose the centroids of the t
wo point setswo point sets Rotate and translate the measured point set with Rotate and translate the measured point set with
respect to the initial poserespect to the initial pose Select those relative positions that yield the miniSelect those relative positions that yield the mini
mum HD valuemum HD value Select the one with the least mean HD.Select the one with the least mean HD.
An image (left) and its distance transform (right)
Test image and Target detected when the contours are superimposed
courtesy of IPAN
Template image
Test image Target detected
courtesy of Cornell Vision Group
ConclusionConclusion Current Progress and Future DirectionsCurrent Progress and Future Directions
TestingTestingDatabaseDatabaseActual Matching with test imagesActual Matching with test images
ReferencesReferences Image and Pattern Analysis Group – Image and Pattern Analysis Group –
http://visual.ipan.sztaki.hu/http://visual.ipan.sztaki.hu/Cornell Computer Vision GroupCornell Computer Vision Group
http://www.cs.cornell.edu/visionhttp://www.cs.cornell.edu/visionRobert M. Haralick, Layne T. Watson, Robert M. Haralick, Layne T. Watson,
Thomas J. Laffey, The Topographic Thomas J. Laffey, The Topographic Primal Sketch. The international Primal Sketch. The international Journal of Robotics Research. Vol. 2, Journal of Robotics Research. Vol. 2, No. 1, Spring No. 1, Spring 19831983