Height estimation of 3D surface based PMS is generally performed in the condition of distance point lighting source and certain ambient light; the intensity of the ambient light is far lower than lighting condition. Due to 3D shape perception of human eyes is more accurate in natural environment. This paper puts forward a new application of using three images to estimate its 3D shape under the natural environment. Experimental results shown that 3D shape estimation of object under the natural environment is accurate, moreover it reduces the complexity of the equipment and the restructured surface is allowed to have certain shading.