EdgeFormer: local patch-based edge detection transformer on point clouds

TL;DR

EdgeFormer is a novel local patch-based transformer network designed for effective edge detection in 3D point clouds, capturing fine details by analyzing local neighborhoods.

Contribution

The paper introduces a two-stage learning-based approach that converts point cloud edge detection into local patch classification, improving detection of fine-grained edges.

Findings

Competitive performance against six baseline methods.

Effective extraction of fine-grained edge details.

Utilizes local patch descriptors for improved accuracy.

Abstract

Edge points on 3D point clouds can clearly convey 3D geometry and surface characteristics, therefore, edge detection is widely used in many vision applications with high industrial and commercial demands. However, the fine-grained edge features are difficult to detect effectively as they are generally densely distributed or exhibit small-scale surface gradients. To address this issue, we present a learning-based edge detection network, named EdgeFormer, which mainly consists of two stages. Based on the observation that spatially neighboring points tend to exhibit high correlation, forming the local underlying surface, we convert the edge detection of the entire point cloud into a point classification based on local patches. Therefore, in the first stage, we construct local patch feature descriptors that describe the local neighborhood around each point. In the second stage, we classify each point by analyzing the local patch feature descriptors generated in the first stage. Due to the conversion of the point cloud into local patches, the proposed method can effectively extract the finer details. The experimental results show that our model demonstrates competitive performance compared to six baselines.