Furniture image marking method, model training method and equipment

Abstract

The embodiment of the invention provides a furniture image marking method, a model training method and equipment. According to the embodiment of the invention, the method comprises the steps: determining a part label to which each spatial point in a 3D furniture model belongs based on point cloud data; converting the 3D furniture model into a 2D furniture image, mapping part labels marked in the 3D furniture model to corresponding pixel points in the 2D furniture image, and recoloring the 2D furniture image based on coloring parameters configured for different part labels so as to generate a part segmentation map corresponding to the 3D furniture model. Therefore, the 2D furniture image corresponding to the 3D furniture model, the component segmentation image and the component label can be used as the training samples to train the furniture component segmentation model, so that the furniture component segmentation model learns knowledge of component segmentation in the 2D furniture image. And the trained furniture part segmentation model can be utilized to realize automatic part-level labeling of massive two-dimensional furniture images.

Description

technical field [0001] The present application relates to the technical field of data processing, in particular to a marking method of furniture images, a model training method and equipment. Background technique [0002] For scenarios such as furniture e-commerce or interior design, it is necessary to provide screening items for furniture products. At present, it usually relies on manual labor to fill in the component attributes of furniture products as the basis for screening items. [0003] However, due to the limitation of labor costs, the filling efficiency is usually relatively low, and this causes a large number of furniture products to lack information on the attributes of these parts, resulting in insufficient screening results under the screening item, affecting the experience of merchants and users. Contents of the invention [0004] Aspects of the present application provide a furniture image marking method, model training method and equipment, to realize comp...

AI Technical Summary

Problems solved by technology

[0003] However, due to the limitation of labor costs, the filling efficiency is usually relatively low, and this causes a large number of furniture products to lack information on the attributes of these parts, resulting in insufficient screening results under the screening item, affecting the experience of merchants and users

Images