Dynamic texture is a part of the natural scene. For example, the landscape of fire, smoke, clouds and wind are dynamic textures. Dynamic texture recognition issue is research hotspot in the field of computer vision and even robot research area. In this paper this it is discussed in detail. First of all, we discuss the significance and status of dynamic texture recognition study. Some methods based on optical flow are introduced. We divide these methods into four kinds: gradient based optical flow calculation recognition method, normal flow based method, optical flow and the level set combination method and optical flow based invariant statistical recognition method. At last we discuss the existing problems and solutions to it. We have an outlook to the prospect of dynamic regional recognition techniques.