PMID- 32450376
OWN - NLM
STAT- PubMed-not-MEDLINE
DCOM- 20200612
LR  - 20200612
IS  - 1879-1026 (Electronic)
IS  - 0048-9697 (Linking)
VI  - 733
DP  - 2020 Sep 1
TI  - Transfer Learning for Crop classification with Cropland Data Layer data (CDL) as 
      training samples.
PG  - 138869
LID - S0048-9697(20)32386-X [pii]
LID - 10.1016/j.scitotenv.2020.138869 [doi]
AB  - Training samples is fundamental for crop mapping from remotely sensed images, but 
      difficult to acquire in many regions through ground survey, causing significant 
      challenge for crop mapping in these regions. In this paper, a transfer learning 
      (TL) workflow is proposed to use the classification model trained in contiguous 
      U.S.A. (CONUS) to identify crop types in other regions. The workflow is based on 
      fact that same crop growing in different regions of world has similar temporal 
      growth pattern. This study selected high confidence pixels across CONUS in the 
      Cropland Data Layer (CDL) and corresponding 30-m 15-day composited NDVI time 
      series generated from harmonized Landat-8 and Sentinel-2 (HLS) data as training 
      samples, trained the Random Forest (RF) classification models and then applied 
      the models to identify crop types in three test regions, namely Hengshui in China 
      (HS), Alberta in Canada (AB), and Nebraska in USA (NE). NDVI time series with 
      different length were used to identify crops, the effect of time-series length on 
      classification accuracies were then evaluated. Furthermore, local training 
      samples in the three test regions were collected and used to identify crops (LO) 
      for comparison. Results showed that overall classification accuracies in HS, AB 
      and NE were 97.79%, 86.45% and 94.86%, respectively, when using TL with NDVI time 
      series of the entire growing season for classification. However, LO could achieve 
      higher classification accuracies earlier than TL. Because the training samples 
      were collected across USA containing multiple growth conditions, it increased the 
      potential that the crop growth environment in test regions could be similar to 
      those of the training samples; but also led to situation that different crops had 
      similar NDVI time series, which caused lower TL classification accuracy in HS at 
      early-season. Generally, this study provides new options for crop classification 
      in regions of training samples shortage.
CI  - Copyright (c) 2020 Elsevier B.V. All rights reserved.
FAU - Hao, Pengyu
AU  - Hao P
AD  - Center for Spatial Information Science and Systems, George Mason University, 
      Fairfax, Virginia, USA. Electronic address: phao@gmu.edu.
FAU - Di, Liping
AU  - Di L
AD  - Center for Spatial Information Science and Systems, George Mason University, 
      Fairfax, Virginia, USA. Electronic address: ldi@gmu.edu.
FAU - Zhang, Chen
AU  - Zhang C
AD  - Center for Spatial Information Science and Systems, George Mason University, 
      Fairfax, Virginia, USA. Electronic address: czhang11@masonlive.gmu.edu.
FAU - Guo, Liying
AU  - Guo L
AD  - Center for Spatial Information Science and Systems, George Mason University, 
      Fairfax, Virginia, USA. Electronic address: lguo2@gmu.edu.
LA  - eng
PT  - Journal Article
DEP - 20200428
PL  - Netherlands
TA  - Sci Total Environ
JT  - The Science of the total environment
JID - 0330500
SB  - IM
OTO - NOTNLM
OT  - Corn
OT  - Cotton
OT  - Cropland Data Layer (CDL)
OT  - Random Forest
OT  - Transfer learning
OT  - USA
COIS- Declaration of competing interest The authors declare that they have no known 
      competing financial interests or personal relationships that could have appeared 
      to influence the work reported in this paper.
EDAT- 2020/05/26 06:00
MHDA- 2020/05/26 06:01
CRDT- 2020/05/26 06:00
PHST- 2020/01/14 00:00 [received]
PHST- 2020/03/28 00:00 [revised]
PHST- 2020/04/19 00:00 [accepted]
PHST- 2020/05/26 06:00 [pubmed]
PHST- 2020/05/26 06:01 [medline]
PHST- 2020/05/26 06:00 [entrez]
AID - S0048-9697(20)32386-X [pii]
AID - 10.1016/j.scitotenv.2020.138869 [doi]
PST - ppublish
SO  - Sci Total Environ. 2020 Sep 1;733:138869. doi: 10.1016/j.scitotenv.2020.138869. 
      Epub 2020 Apr 28.