Model structure, model training method, monomer method, equipment and medium

The invention discloses a model structure, a model training method, a monomer method, equipment and a medium. The model training method comprises the following steps: acquiring original three-dimensional point cloud data of a large-scene ground feature; making the original three-dimensional point cloud data into a standard sample format file; preprocessing the point cloud sample in the standard sample format file to generate a PKL format sample file; the method comprises the following steps: constructing a large-scene ground feature monomerization model, wherein the large-scene ground feature monomerization model comprises a coding module, a backbone network, a target generation module, a feature fusion module, a Point-RoIAlign module and an instance prediction network; and training the large-scale scene ground feature monomerization model by using the point cloud sample in the PKL format sample file to obtain a trained large-scale scene ground feature monomerization model. According to the method, prediction