激光与光电子学进展, 2018, 55 (2): 022802, 网络出版: 2018-09-10   

结合均值漂移分割与全卷积神经网络的高分辨遥感影像分类 下载: 1433次

High Resolution Remote Sensing Image Classification Combining with Mean-Shift Segmentation and Fully Convolution Neural Network
作者单位
1 中国矿业大学环境与测绘学院, 江苏 徐州 221116
2 国家测绘地理信息局卫星测绘应用中心, 北京 100048
摘要
针对目前遥感影像分类应用中常用的浅层机器学习算法无法满足当前海量遥感影像数据环境下分类精度的问题,提出了一种将全卷积神经网络应用于遥感影像分类的方法;为了减少影像特征图在池化过程中自身特征的丢失,增加池化层与反卷积层的融合;为了提高融合的可靠性,增加尺度变换层;为了获得更精细的边缘分类结果,考虑像素之间的空间相关性,采用均值漂移聚类分割获取像素的空间关系,通过统计聚类区域像素概率的和最大、方差最小的方法确定该区域对象的类别;选取典型地区的影像进行分类实验,并将所提出的分类方法与全卷积神经网络、支持向量机、人工神经网络方法进行对比。结果表明,所提出的分类方法的精度明显高于传统机器学习方法的精度。
Abstract
Aim

ing at the problem that the shallow machine learning algorithm commonly used in remote sensing image classification application cannot satisfy the classification accuracy in the current mass remote sensing image data environment, we propose a method to apply the fully convolution neural network to the remote sensing image classification. To reduce the loss of image feature map in the pooling process, we add the fusion of the pool layer and the deconvolution layer. To improve the reliability of fusion, we add the scale layer. To obtain finer edge classification results, considering the spatial correlation between pixels mean-shift clustering is used to obtain the spatial relationship of pixels. Classes of regional objects are determined by the maximum sum and the minimum variance of the regional pixel probabilities. Images of typical regions are chosen to carry out the classification experiments, and the classification method proposed in this paper is compared with those of the fully convolution neural network, support vector machine, and artificial neural network. The results show that the accuracy of the classification method proposed in this paper is obviously higher than that of the traditional machine learning methods.

方旭, 王光辉, 杨化超, 刘慧杰, 闫立波. 结合均值漂移分割与全卷积神经网络的高分辨遥感影像分类[J]. 激光与光电子学进展, 2018, 55(2): 022802. Xu Fang, Guanghui Wang, Huachao Yang, Huijie Liu, Libo Yan. High Resolution Remote Sensing Image Classification Combining with Mean-Shift Segmentation and Fully Convolution Neural Network[J]. Laser & Optoelectronics Progress, 2018, 55(2): 022802.

本文已被 10 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!