Provided are a method, a storage medium and an apparatus for converting a 2D picture set to a 3D model. The method includes: identifying the category of each object contained in each picture in the 2D picture set and outline information of each object by using a deep learning algorithm; extracting detail characteristic information of each object in each picture in the 2D picture set by using a computer vision algorithm, wherein the detail feature information at least includes texture information, color feature information and shape feature information of each object; matching the 3D model based on the category, the shape feature information and the outline information of each object, wherein the 3D model matched successfully is the 3D model of the object; and mapping the texture information and the color feature information of each object onto the 3D model of each object.