Point clouds denoising is one of surface reconstruction’s key technologies. This paper makes full use of space unorganized points’ inner information, to re-interpret over-domain average method of image denois- ing theory, so that the theory of two-dimensional space can be applied to three-dimensional space’s point data. And based on the space analytic geometry theory, it creates a method of three-dimensional space’s unorganized point clouds denoising. The experimental results show that this method better retains detail fea- tures of the unorganized points of 3D shape and noise is obviously suppressed after point clouds denoising.