Autonomous　navigation　is　a　key　part　for　soft-landing　asteroid,　and　the　technology　of　feature　recognition　and　matching　is　critical　in　this　part.　Some　approaches　were　discussed　in　this　paper.　First,　we　use　two-dimensional　maximum　entropy　thresholding　segmentation　for　extraction　the　extract　features,　and　then　apply　Fourier　descriptors　for　feature　matching.　Combing　Fourier　discriptiors　with　PCA,　and　with　the　help　of　vector　relationship　of　features,　we　conduct　a　series　of　feature　matching　experiments.　The　experimental　results　show　that　this　method　can　extract　and　match　features　effectively.