《内容提要499教学文稿》由会员分享，可在线阅读，更多相关《内容提要499教学文稿（46页珍藏版）》请在金锄头文库上搜索。
1、LANL/SLAAC DARPA CollaborationSLAAC Team Meeting 3-99LANL Challenge ProblemsKevin McCabekmccabelanl.gov505-667-0728Multi-DimensionalImaging Jeff BlochUltra-Wideband Coherent RF Mark DunhamLANL/SLAAC DARPA CollaborationDAPSMultiDimensionalMultiDimensionalImage ProcessingImage ProcessingUltra-Wide Ban
2、dUltra-Wide BandRadio FrequencyRadio FrequencyReal Time Processing of Multi/Hyper Spectral or Time domain Data CubesReal Time Processing wide band RF dataReConfigurable Computing Hardware:IP51/RCA-2/DARPA/(RCA-3)/CommercialRCC Architecture Development/Deployment Hardware/Software EnvironmentClassifi
3、cation/Compression/RecognitionAlgorithms for fixed point RCC HardwareALDEBARAN Mark Dunham Bellatrix Scott Robinson Capella R. Dingler Cibola Mike CaffreyCALIOPE Kurt MooreHIRIS/MTI John Szymanski James TheilerSHSRULLIHyperspectralDemonstrationsDARPA Collaboration:RCC HW/SWTool EvaluationMike Caffre
4、y, Phil Blain, Noor Khalsa, Tony Rose, Tony NelsonMike Caffrey, Tony Salazaar, John Layne, Jan Friego John Szymanski, James Theiler, Jeff Bloch, Kurt Moore, Chris BrislawnSteven Brumby Reid Porter,Simon PerkinsKevin McCabeKurt MooreUniversity Collaborations: BYU, UT,.DII: “Rapid Feature Identificati
5、onUsing RCC Technology andGenetic Algorithms”Jeff Bloch, John SzymanskiFORTEV-SENSORLANL/SLAAC DARPA CollaborationCollaboration Phase 1Represent Challenge ProblemsUltra-Wide Band RF (UWBRF) Signal ProcessingMulti-Dimensional Image Processing (MDIP)Provide specific challenge problem descriptions to A
6、CS investigatorsMDIP: http:/nis-www.lanl.gov/nis-projects/daps/UWBRF: http:/www.lanl.gov/rcc/Seek out and collaborate with ACS investigators whose work matches our needISI - DRP and RRPNorthwestern - Matlab compilerPtolemy - System level analysis toolValidate Hardware or Software StrategiesLANL/SLAA
7、C DARPA CollaborationLANL/SLAAC DARPA CollaborationRFIPJeff Bloch, 505-665-2568, jblochlanl.govJohn Szymanski, szymanskilanl.gov, 505-665-9371James Theiler, jtheilerlanl.gov, 505-665-5682LANL/SLAAC DARPA CollaborationRFIPObjectiveManipulate image processing steps carried out on RCC hardware to devel
8、op remote sensing algorithms for classifying and identifying features of interest to an image analyst.Provide software suite and hardware to work in host workstation(s)Search speeds comparable to archive retrieval timesPlatform and RCC hardware independenceScalable within a platform and via networke
9、d platformsApproachRapid evolution of feature recognition procedure via:Hardware accelerator containing tunable image processing operatorsSoftware engine that parallelizes a search and manipulates accelerated operators to maximize performance against truth dataLANL/SLAAC DARPA CollaborationRFIPCurre
10、nt PlanDevelop non-real time demonstration (proof of concept) of an algorithm to evolve image classification procedures for identifying features of interest (fully funded by RFIP)Select image operators amenable to RCC accelerationSelect algorithm framework and softwareSelect simulation environment (
11、IDL, Perl, etc.)Select test datasetsDevelop, write, and execute an all software demonstrationDemonstrate ability of RCC to dramatically speed up image processing steps (RFIP - SLAAC partnership)Select demonstration RCC hardware (SLAAC-1 insertion opportunity)Develop tunable operator architecture in
12、VHDLSelect some image operators and code them in VHDLDevelop software engine to drive a single RCCBenchmark accelerated operators against software operatorsLANL/SLAAC DARPA CollaborationRFIPFuture Plan (2 DII proposals being submitted this month)Fully develop RCC accelerated workstationAdd to select
13、ion image operators amenable to RCC acceleration and code them in VHDLRefine algorithm framework and softwareDefine and procure RCC computer suitable for analysts workstationBroaden test datasetsBenchmark against all software solutionDemonstrate parallelizability and scalability of approach on multi
14、ple workstationsImplement prior all software solution across multiple workstationsDevelop a parallel execution schemeAccelerated algorithm evolution against one truth data setAccelerated processing of search dataDevelop advanced user interfaceBenchmark against single workstation all software solutio
15、nLANL/SLAAC DARPA CollaborationRFIPRFIP Project StatusNon-real time demonstration functional Ability to demonstrate a limited number of operatorsFurther refinement of tunable operators approach ongoingRapid evolution of a Water Finding Procedure demonstratedBenchmark of all software solution to be d
16、oneCurrent RFIP funding ends 12/99RFIP SLAAC collaboration effortDemonstration of RCCCandidate image operators selectedTunable operator architecture concept under developmentTargeting of RCC hardware to be doneDevelop software engine to drive a single RCC to be doneBenchmark against software operators to be doneLimited demonstration to prove principle by 12/99LANL/SLAAC DARPA CollaborationSpectralSpectralSpectralSpatialSpatialSpectralSpectralSpectralSpectralSpatialFeature 1Feature 3Feature 2Thre
