With　the　development　of　computer　vision　technology,　image　classification　has　become　an　important　research　field.　In　order　to　achieve　automated　classification　of　Tang　Ka　images,　this　paper　proposes　a　visual　feature　based　classification　method　for　Tang　Ka　images.　The　method　first　uses　HOG　features,　Local　binary　patterns　and　color　moments　to　extract　image　features,　and　then　uses　support　vector　machine　classifier　to　classify　images.　The　data　in　this　article　shows　that　the　highest　accuracy　rate　of　the　Hantangka　obtained　through　support　vector　machine　classification　and　filtering　is　96.5%.　The　experimental　results　show　that　this　method　performs　well　in　the　classification　of　heritage　thangka　images　and　has　a　high　classification　accuracy.　　©　2023　IEEE.