Adaptive Anchor Networks for Multi-Scale Object Detection in Remote Sensing Images

Miaohui Zhang,Jun Wang,Yunzhong Chen,Bingxue Lv,Xianxing Liu

doi:10.1109/access.2020.2982658

Miaohui Zhang, Jun Wang + Show 3 more

Open Access

https://doi.org/10.1109/access.2020.2982658

Copy DOI

Abstract

Accurate and effective object detection in remote sensing images plays an extremely important role in marine transport, environmental monitoring and military operations. Due to the powerful ability of feature representation, region-based convolutional neural networks (RCNNs) have been widely used in this field, which firstly generate candidate regions through extracted feature maps and then classify and locate objects. However, most of existing methods generally use traditional backbone networks to extract feature maps with a decreased spatial resolution because of the continuous down-sampling, which will weaken the information detected from small objects. Besides, sliding windows strategy is employed in these methods to generate fixed anchors with a preset scale on feature maps, which is inappropriate for multi-scale object detection in remote sensing images. To solve the above problems, a novel and effective object detection framework named DetNet-FPN (Feature Pyramid Network) is proposed in this paper, in which a feature pyramid with strong feature representation is created by combining feature maps of different spatial resolution, at the same time, the resolution of feature maps is maintained by involving dilation convolutions. Furthermore, to match the proposed backbone, the GA (Guided Anchoring)-RPN strategy is improved for adaptive anchor generation, this strategy simultaneously predicts the locations where the center of objects are likely to exist as well as the scales and aspect ratios at different locations. Extensive experiments and comprehensive evaluations demonstrate the effectiveness of the proposed framework on DOTA and NWPU VHR-10 datasets.

Highlights

With the rapid development of remote sensing technologies, the number of remote sensing images has been growing dramatically, and object detection in remote sensing images is attracting more and more attention owing to its wide application, such as urban planning, environmental monitoring and precision agriculture
The performance of object detection has been significantly improved due to the rapid progress of deep convolutional neural networks
A lot of region proposals at various scales are generated on the fused multi-scale feature maps using GA-Region Proposal Network (RPN), different from the sliding window methods, GA-RPN can generate adaptive anchors matching with the shapes of objects, which is conducive to the accurate regression for objects at different scales

Summary

Introduction

With the rapid development of remote sensing technologies, the number of remote sensing images has been growing dramatically, and object detection in remote sensing images is attracting more and more attention owing to its wide application, such as urban planning, environmental monitoring and precision agriculture. Remote sensing images have the following characteristics:. (1) Objects with various scale and aspect ratios. Remote sensing images are commonly shot from high altitudes, ranging from hundreds to thousands of meters, and the resolution of the sensor varies obviously, so the change of object scale in remote sensing images is more obvious than that in natural scene. For different airplanes (Fig. 1(a, b)), object sizes may vary greatly due to the changes in shooting heights, and an airplane may take up only a small proportion in a picture, but a ground track field may occupy the biggest part of the total image (Fig. 1(a, c)). There are a significant difference in aspect ratios between a bridge/harbor and an airplane (Fig. 1(d/f and a))

Methods

Results

Conclusion