PointPPE: A precise recognition method for complex machining features based on point cloud analysis network with polynomial positional encoding

Guiyu Jiang

doi:10.1016/j.displa.2025.103214

PointPPE: A precise recognition method for complex machining features based on point cloud analysis network with polynomial positional encoding

2025-09-07

0

PRE

AI

DOI：10.1016/j.displa.2025.103214

Original

Original request for help

Save

Abstract

En 中文

• A high-fidelity dataset of machining feature point clouds is established. • The position encoding module is enhanced by multiple polynomial functions. • Local feature analysis is improved through a channel attention context fusion module. • A feature recognition network is developed for learning complex machining features.

Keywords:

Machining feature recognition

Point cloud

Positional encoding

Part segmentation

PointPPE: A precise recognition method for complex machining features based on point cloud analysis network with polynomial positional encoding

Abstract

Journal

Organization