Fusion of Deep Convolutional Neural Networks for Land Cover Classification of High-Resolution Imagery
Fusion of Deep Convolutional Neural Networks for Land Cover Classification of High-Resolution Imagery
复制标题
DOI：
10.1109/lgrs.2017.2722988
复制
发表时间：
2017-09-01
影响因子：
4.8
通讯作者：
Nivin, Tyler W.
中科院分区：
文献类型：
--
作者：
Scott, Grant J.;Marcum, Richard A.;Nivin, Tyler W.
深度卷积神经网络（DCNN）最近成为许多图像分类应用的最高性能方法，包括高分辨率遥感图像的自动土地覆盖分类。在这封信中，我们研究了各种融合技术，将多个DCNN土地覆盖分类器融合到一个聚合分类器中。虽然特征级融合广泛用于深度神经网络，但我们的方法专注于分类/信息级的融合。在这里，我们训练了三种不同的DCNN：CaffeNet，GoogLeNet和ResNet50。各种信息融合方法，包括投票，加权平均，模糊积分，然后评估的有效性。特别是，我们使用DCNN交叉验证结果来确定模糊积分的输入密度，然后进行进化优化。这种新的方法产生的最先进的分类结果高达99.3%的UC默塞德数据集和99.2%的RSD数据集。
Deep convolutional neural networks (DCNNs) have recently emerged as the highest performing approach for a number of image classification applications, including automated land cover classification of high-resolution remote-sensing imagery. In this letter, we investigate a variety of fusion techniques to blend multiple DCNN land cover classifiers into a single aggregate classifier. While feature-level fusion is widely used with deep neural networks, our approach instead focuses on fusion at the classification/information level. Herein, we train three different DCNNs: CaffeNet, GoogLeNet, and ResNet50. The effectiveness of various information fusion methods, including voting, weighted averages, and fuzzy integrals, is then evaluated. In particular, we used DCNN cross-validation results for the input densities of fuzzy integrals followed by evolutionary optimization. This novel approach produces the state-of-the-art classification results up to 99.3% for the UC Merced data set and the 99.2% for the RSD data set.