Improved Winter Wheat Spatial Distribution Extraction Using A Convolutional Neural Network and Partly Connected Conditional Random Field

被引：12

作者：

Wang, Shouyi ^{[1
]}

Xu, Zhigang ^{[2
]}

Zhang, Chengming ^{[1
]}

Zhang, Jinghan ^{[3
]}

Mu, Zhongshan ^{[4
]}

Zhao, Tianyu ^{[4
]}

Wang, Yuanyuan ^{[1
,5
]}

Gao, Shuai ^{[6
]}

Yin, Hao ^{[1
]}

Zhang, Ziyun ^{[1
]}

机构：

[1] Shandong Agr Univ, Coll Informat Sci & Engn, 61 Daizong Rd, Tai An 271000, Shandong, Peoples R China

[2] Hubei Univ Technol, Sch Comp Sci, 28 Nanli Rd, Wuhan 430068, Hubei, Peoples R China

[3] Shandong Taian Middle Sch, 6 Hushandong Rd, Tai An 271000, Shandong, Peoples R China

[4] South To North Water Transfer East Route Shandong, Jinan 250000, Shandong, Peoples R China

[5] Shandong Technol & Engn Ctr Digital Agr, 61 Daizong Rd, Tai An 271000, Shandong, Peoples R China

[6] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, 9 Dengzhuangnan Rd, Beijing 100094, Peoples R China

来源：

REMOTE SENSING | 2020年 / 12卷 / 05期

基金：

国家重点研发计划;

关键词：

convolutional neural network; partly connected conditional random field; remote sensing imagery; image segmentation; refined edge; prior knowledge; winter wheat; Gaofen-2; image; Tai'an; China; TIME-SERIES; BOUNDARY CONSTRAINT; OBJECT DETECTION; LAND-COVER; CLASSIFICATION; REGION; MODEL; SVM;

D O I：

10.3390/rs12050821

中图分类号：

X [环境科学、安全科学];

学科分类号：

08 ; 0830 ;

摘要：

Improving the accuracy of edge pixel classification is crucial for extracting the winter wheat spatial distribution from remote sensing imagery using convolutional neural networks (CNNs). In this study, we proposed an approach using a partly connected conditional random field model (PCCRF) to refine the classification results of RefineNet, named RefineNet-PCCRF. First, we used an improved RefineNet model to initially segment remote sensing images, followed by obtaining the category probability vectors for each pixel and initial pixel-by-pixel classification result. Second, using manual labels as references, we performed a statistical analysis on the results to select pixels that required optimization. Third, based on prior knowledge, we redefined the pairwise potential energy, used a linear model to connect different levels of potential energies, and used only pixel pairs associated with the selected pixels to build the PCCRF. The trained PCCRF was then used to refine the initial pixel-by-pixel classification result. We used 37 Gaofen-2 images obtained from 2018 to 2019 of a representative Chinese winter wheat region (Tai'an City, China) to create the dataset, employed SegNet and RefineNet as the standard CNNs, and a fully connected conditional random field as the refinement methods to conduct comparison experiments. The RefineNet-PCCRF's accuracy (94.51%), precision (92.39%), recall (90.98%), and F1-Score (91.68%) were clearly superior than the methods used for comparison. The results also show that the RefineNet-PCCRF improved the accuracy of large-scale winter wheat extraction results using remote sensing imagery.

引用

页数：25