Abstract: Nowadays,starpatternrecognitionmethodsusedonstarsensorshaveshortcomingsofredundancymatch,matchingslow,lowrecognitionrate，andsoon.Toovercomethis，thestarpatternrecognitionalgorithmwasstudied，andanewalgorithmwasproposed.Firstly,redundancyinformationfromtraditionalstarcatalogwasextracted.Theskycoveragewasdividedtomakesurethatdifferentskycoveragehasthesimilarnavstarnumber.Secondly，featureinformationfromthephotowasextractedtomatchthefeatureinformationdatabase.Afterthealgorithmwasdesigned,thestarsensorswiththefieldofview15°×15°wereselectedandsimulated.Simulationresultsshowthatthematchingofthewholeskycanbefinishedin15sbythisalgorithm,andthesuccessratecanreachover98%.

Key words: stallermapidentification, stallertablereconstruction, imageprocessing, characterizedgraphicsmatching, starsensor, wholeskycoverage