[1]徐风,孙万砚.基于非对称全局卷积神经网络的遥感图像识别方法[J].林业工程学报,2020,5(06):137-142.[doi:10.13360/j.issn.2096-1359.202001037]
 XU Feng,SUN Wanyan.Remote sensing scene recognition using unsymmetrical nonlocal convolutional neural network[J].Journal of Forestry Engineering,2020,5(06):137-142.[doi:10.13360/j.issn.2096-1359.202001037]
点击复制

基于非对称全局卷积神经网络的遥感图像识别方法()
分享到:

《林业工程学报》[ISSN:2096-1359/CN:32-1862/S]

卷:
5
期数:
2020年06期
页码:
137-142
栏目:
装备与信息化
出版日期:
2020-11-01

文章信息/Info

Title:
Remote sensing scene recognition using unsymmetrical nonlocal convolutional neural network
文章编号:
2096-1359(2020)06-0137-06
作者:
徐风孙万砚
南京林业大学信息科学技术学院,南京210037
Author(s):
XU Feng SUN Wanyan
College of Information Sciences and Technology, Nanjing Forestry University, Nanjing 210037, China
关键词:
卷积神经网络遥感图像图像识别智能林业
Keywords:
convolutional neural network remote sensing scene image recognition intelligent forestry
分类号:
TP751.1
DOI:
10.13360/j.issn.2096-1359.202001037
文献标志码:
A
摘要:
全局卷积神经网络(Nonlocal CNN)是一种增强的深度卷积神经网络模型,在图像处理任务中用来获取输入图像的全局上下文信息。考虑到部分种类遥感图像内部存在重复像素块或重复目标的特性,笔者提出了一种非对称的全局卷积结构(UNC),用于减少传统全局卷积结构(NC)的计算量,从而提升模型的效率。首先对NC结构中键值对(Keyvalue)对应的特征图通过下采样降低分辨率;然后将查询值(Query)和下采样后的Keyvalue特征图逐像素相乘,以获取输入图像的全局上下文表示和加权特征;最后将加权特征和残差连接相加,得到UNC结构的输出。和传统的NC结构相比,UNC是一种更高效、通用的全局上下文信息获取结构,可以用于多种深度CNN模型中。为了验证UNC的有效性,在4种公开的遥感图像目标识别数据集上进行了实验,结果表明:相比NC结构,UNC可以最多减少24.53%的参数以及整体49.1%的浮点运算量,并且在RSSCN7、UCML、WHURS19和AID可分别最高取得96.63%,99.16%,98.90%和96.28%的识别准确率,具有较强的实用性和普适性。
Abstract:
Compared with traditional shallow learning methods, deep learning based methods have the advantages of high efficiency, outstanding performance and strong expression ability. Among these methods, the image processing technology based on the convolutional neural network is widely used in the fields of image classification, semantic segmentation and object detection. In particular, Nonlocal convolutional neural network (Nonlocal CNN) is an enhanced deep convolutional neural network model, which has attracted much attention in recent years. In image processing tasks, the Nonlocal CNN can be used to obtain the global contextual rather than the limited local information of an input image, which is proven to be beneficial for image recognition tasks. In a large number of computer vision tasks, the effectiveness and feasibility of the Nonlocal CNN have been proved. Considering the fact that there are numerous repeated pixelpatches and objects in some categories of remote sensing images (i.e., forest, farmland and beach), the repeated calculations for the same pixelpatches not only would reduce the model efficiency but also may worsen the performance of the model. To address this problem, this study proposed an unsymmetrical nonlocal convolution (UNC) module to reduce the calculation time and space complexity of the Nonlocal CNN, so that the model efficiency can be improved. Specifically, a downsampling operation on feature maps of Keyvalue pairs in the classical nonlocal convolutional module was performed firstly; then, a global object context estimation and weighted feature maps of the input image was obtained by the pixelwise multiplication of Query and Keyvalue pair; finally, the residual connection and weighted feature maps were added together as the output for the UNC module. Compared with the classical Nonlocal CNN, the UNC has the same implementation process as the Nonlocal CNN, but the proposed UNC was an efficient and general context acquisition module that can be used in a variety of deep learning CNN models. In order to demonstrate its effectiveness, experiments were carried out on four public remote sensing object recognition datasets. The experimental results showed that UNC could reduce the model space complexity by up to 24.53% as well as 49.1% on the floating point operations, and obtain the recognition accuracies of 96.63%, 99.16%, 98.90% and 96.28% in RSSCN7, UCML, WHURS19, and AID, respectively, which demonstrated that UNC had a strong practicality and universality.

相似文献/References:

[1]许等平,任怡,闫哲,等.基于CNN的无人机遥感影像质量评价[J].林业工程学报,2018,3(05):121.[doi:10.13360/j.issn.2096-1359.2018.05.019]
 XU Dengping,REN Yi,YAN Zhe,et al.The CNN driven quality evaluation for UAV remote sensing images[J].Journal of Forestry Engineering,2018,3(06):121.[doi:10.13360/j.issn.2096-1359.2018.05.019]
[2]徐风,苗哲,业巧林.基于卷积注意力模块的端到端遥感图像分类[J].林业工程学报,2020,5(04):133.[doi:10.13360/j.issn.2096-1359.201907003]
 XU Feng,MIAO Zhe,YE Qiaolin.End-to-end remote sensing image classification framework based on convolutional block attention module[J].Journal of Forestry Engineering,2020,5(06):133.[doi:10.13360/j.issn.2096-1359.201907003]

备注/Memo

备注/Memo:
收稿日期:2020-01-29? ? ?修回日期:2020-03-29
基金项目:江苏省自然科学基金(BK20171453)。
作者简介:徐风,男,实验师,研究方向为图像处理与机器学习。E-mail: fufeng_njfu@163.com
更新日期/Last Update: 2020-11-11