AUTHOR=Liu Zhe , Zhang Lin , Yu Yaoqi , Xi Xiaojie , Ren Tianwei , Zhao Yuanyuan , Zhu Dehai , Zhu A-xing 

TITLE=Cross-Year Reuse of Historical Samples for Crop Mapping Based on Environmental Similarity

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 12 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2021.761148

DOI=10.3389/fpls.2021.761148

ISSN=1664-462X

ABSTRACT=Crop classification maps are fundamental data for global change research, regional agricultural regulation, fine production and insurance services. The key to crop classification is samples, but it is very time-consuming in annual field sampling. Therefore, how to use historical samples on crop classification for future years at a lower cost is a research hotspot. By constructing the spectral feature vector of each historical sample pixel and the neighboring pixels of the target year, we produced new samples and classified in the target year. Specifically, based on the environmental similarity, we first calculated similarities of every two pixels between each historical year and target year, and took the neighboring pixel with the highest local similarity as the potential samples. Then, cluster analysis was performed on those potential samples of the same crop, and the class with more pixels is selected as newly generated samples for classifying of target year. The experiment on Heilongjiang Province, China showed that this method can generate new samples with a uniform spatial distribution, and the proportion of various crops is consistent with the field data in historical years. The overall accuracy of the target year by the newly generated sample and the real sample is 61.57% and 80.58% respectively. The spatial pattern of maps obtained by the two models is basically the same, and the classification based on the newly generated samples identified rice better. This method overcomes the problem of insufficient samples caused by the difficulties on visual interpretation and high cost on field sampling, effectively improves the utilization rate of historical samples, and provides a new idea for crop mapping in areas lacking field samples of the target year.