Method and device for creating physical scene model, and method and device for aided driving

A method and device for creating a model of a physical scene including a variety of objects, and a method and a device for aided driving are provided. The method for creating a model of a physical scene including a variety of objects comprises the following steps: acquiring an image of the physical scene; extracting the features of each of the objects in the image; judging whether there is a scene model; when judging that there is no scene model, carrying out scene model initializing based on the extracted features of each object to get an initialized scene model; and outputting the initialized scene model.