An Improved BoxInst Model for Plane Instance Segmentation in Remote Sensing Images

Shangzheng Jiang, Qingzhong Jia, Fengkun Luo, Tao Yang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

When traditional instance segmentation model is applied to the plane target in the remote sensing image, high-frequency co-occurrence background is easy to appear, which affects the segmentation precision. To solve this problem, this paper proposes an improved BoxInst model for the plane segmentation in remote sensing images. The improved model suppress high-frequency co-occurrence background by introducing background constraint loss. Foreground constraint loss is proposed to avoid trivial solutions caused by the background constraint loss. The improved model is verified on the plane data set, and the average precision of the proposed model is improved by 6% compared with the original BoxInst model.

Original languageEnglish
Title of host publicationIMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference
EditorsBing Xu
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1694-1699
Number of pages6
ISBN (Electronic)9781728185347
DOIs
Publication statusPublished - 18 Jun 2021
Event4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021 - Chongqing, China
Duration: 18 Jun 202120 Jun 2021

Publication series

NameIMCEC 2021 - IEEE 4th Advanced Information Management, Communicates, Electronic and Automation Control Conference
ISSN (Print)2693-2814
ISSN (Electronic)2693-2776

Conference

Conference4th IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference, IMCEC 2021
Country/TerritoryChina
CityChongqing
Period18/06/2120/06/21

Keywords

  • BoxInst
  • constraint loss
  • instance segmentation
  • remote sensing image

Fingerprint

Dive into the research topics of 'An Improved BoxInst Model for Plane Instance Segmentation in Remote Sensing Images'. Together they form a unique fingerprint.

Cite this