ECFRNet: Effective corner feature representations network for image corner detection

作者：

Highlights：

• The existing machine learning-based interest point detection methods are reviewed.

• A novel network is designed to extract multi-scale corner feature.

• A new multi-scale corner detection mechanism with a novel loss function is proposed.

• The designed architecture has considered how to detect corners more accurately.

摘要

•The existing machine learning-based interest point detection methods are reviewed.•A novel network is designed to extract multi-scale corner feature.•A new multi-scale corner detection mechanism with a novel loss function is proposed.•The designed architecture has considered how to detect corners more accurately.

论文关键词：Corner detection,Characteristics of interest points (corners and blobs),Effective corner feature representations network

论文评审过程：Received 23 May 2022, Revised 8 August 2022, Accepted 21 August 2022, Available online 27 August 2022, Version of Record 5 September 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118673