A New Spatial-Oriented Object Detection Framework for Remote Sensing Images

doi:10.1109/TGRS.2021.3127232

2022-01-01

PRE

Dawen Yu

Shunping Ji *

DOI：10.1109/TGRS.2021.3127232

原文链接

原文求助

摘要

En 中文

Although the orientation and scale properties of the objects in remote sensing images have been widely considered in the modern deep learning-based object detection methods, the spatial distribution property of objects has rarely been investigated. There is a distinct spatial distribution difference between close-range objects and remote sensing objects: the former may exhibit extensive mutual occlusion and overlap, whereas the latter rarely overlap. A current remote sensing object detection algorithm that ignores the spatial distribution difference may unnecessarily apply the massive anchor-based proposal bounding box generation and nonmaximum suppression (NMS) operations. In this article, considering the unique spatial distribution of remote sensing objects, and also the other spatial properties, we propose a novel, compact, and spatial-oriented object detection framework for remote sensing images. The proposed two-stage convolutional neural network (CNN) framework, which we call the Remote-sensing Spatial Adaptation DETector (RSADet), considers the spatial distribution, scale, and orientation/shape varieties of the objects in remote sensing images. In the first stage, each object instance is inferred on the scale-attention boosted CNN heatmaps to generate candidate bounding boxes, instead of using the anchor-based proposal box generation and NMS. In the second stage, deformable convolutions are introduced to adapt to the geometric variations of different object instances and to avoid the impact of complex and changeable backgrounds. A new bounding box confidence (IoU score) prediction branch is introduced as a convenient constraint for eliminating unreliable boxes and improving performance. Experiments were conducted on a large single-class remote sensing object detection dataset (the Ningbo Pylon dataset) built as part of this study and an open-source extraordinarily large multiclass dataset (the object DetectIon in Optical Remote sensing image (DIOR) dataset). Compared with the advanced detectors from both the computer vision and remote sensing communities, the proposed RSADet achieved state-of-the-art performance on both datasets.

Keyword:

Remote sensing

Object detection

Feature extraction

Proposals

Graphical models

Distribution functions

Detectors

Convolutional neural network (CNN)

object detection

object spatial distribution

object-based proposal box inference

remote sensing images

期刊

IEEE Transactions on Geoscience and Remote Sensing

IF:

8.6

论文数:

2.1W

被引数:

10.7W

机构

wuhan university

学者数:

8.1W

论文数: 5.8W

被引数: 70

引用论文

Blastococcus capsensis sp. nov., isolated from an archaeological Roman pool and emended description of the genus Blastococcus, B. aggregatus, B. saxobsidens, B. jejuensis and B. endophyticus

International Journal of Systematic and Evolutionary Microbiology

IF0

2016-11-01

OAAI

Karima Hezbri; Moussa Louati; Imen Nouioui; Maher Gtari; Manfred Rohde; Cathrin Spröer; Peter Schumann; Hans-Peter Klenk; Faten Ghodhbane-Gtari; Maria del Carmen Montero-Calasanz

Proceedings of the 2018 10th International Conference on Computer and Automation Engineering

IF0

2018-02-24

PREAI

Koji Abe; Hiroyuki Morita; Takahiro Hayashi

Object Detection in High Resolution Remote Sensing Imagery Based on Convolutional Neural Networks With Suitable Object Scale Features 基于合适目标尺度特征的卷积神经网络高分辨率遥感影像目标检测

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING

IF8.6

2020-03-01

103

PREAI

Dong, Zhipeng; Wang, Mi; Wang, Yanli; Zhu, Ying; Zhang, Zhiqi