文章基本信息

标题：An Elliptic Centerness for Object Instance Segmentation in Aerial Images
本地全文：下载
作者：Yixin Luo ; Jiaming Han ; Zhou Liu 等
期刊名称：Journal of Remote Sensing
电子版ISSN：2694-1589
出版年度：2022
卷号：2022
DOI：10.34133/2022/9809505
语种：English
出版社：American Association for the Advancement of Science (AAAS)
摘要：Instance segmentation in aerial images is an important and challenging task. Most of the existing methods have adapted instance segmentation algorithms developed for natural images to aerial images. However, these methods easily suffer from performance degradation in aerial images, due to the scale variations, large aspect ratios, and arbitrary orientations of instances caused by the bird’s-eye view of aerial images. To address this issue, we propose an elliptic centerness (EC) for instance segmentation in aerial images, which can assign the proper centerness values to the intricate aerial instances and thus mitigate the performance degradation. Specifically, we introduce ellipses to fit the various contours of aerial instances and measure these fitted ellipses by two-dimensional anisotropic Gaussian distribution. Armed with EC, we develop a one-stage aerial instance segmentation network. Extensive experiments on a commonly used dataset, the instance segmentation in aerial images dataset (iSAID), demonstrate that our proposed method can achieve a remarkable performance of instance segmentation while introducing negligible computational cost.