Deep learning based multi-temporal crop classification

By Liheng Zhong, Lina Hu, Hang Zhou. Remote Sensing of Environment, 2019

894 citations recorded by the source corpus.

Abstract This study aims to develop a deep learning based classification framework for remotely sensed time series. The experiment was carried out in Yolo County, California, which has a very diverse irrigated agricultural system dominated by economic crops. For the challenging task of classifying summer crops using Landsat Enhanced Vegetation Index (EVI) time series, two types of deep learning models were designed: one is based on Long Short-Term Memory (LSTM), and the other is based on one-dimensional convolutional (Conv1D) layers. Three widely-used classifiers were also tested for comparison, including a gradient boosting machine called XGBoost, Random Forest, and Support Vector Machine. Although LSTM is widely used for sequential data representation, in this study its accuracy (82.41%) and F1 score (0.67) were the lowest among all the classifiers. Among non-deep-learning classifiers, XGBoost achieved the best result with 84.17% accuracy and an F1 score of 0.69. The highest accuracy (85.54%) and F1 score (0.73) were achieved by the Conv1D-based model, which mainly consists of a stack of Conv1D layers and an inception module. The behavior of the Conv1D-based model was inspected by visualizing the activation on different layers. The model employs EVI time series by examining shapes at various scales in a hierarchical manner. Lower Conv1D layers of the optimized model capture small scale temporal variations, while upper layers focus on overall seasonal patterns. Conv1D layers were used as an embedded multi-level feature extractor in the classification model which automatically extracts features from input time series during training. The automated feature extraction reduces the dependency on manual feature engineering and pre-defined equations of crop growing cycles. This study shows that the Conv1D-based deep learning framework provides an effective and efficient method of time series representation in multi-temporal classification tasks.

Subfields: Ecology, Plant Science

View Paper: https://www.semanticscholar.org/paper/efa0048d1f326a369df9c704118a54e062dab337

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecology

In Plant Science

Citations in the corpus

County-Level Soybean Yield Prediction Using Deep CNN-LSTM Model (2019, methodology): “Remote sensing has been widely used in agricultural applications including cropland cover classification, drought stress estimation, and yield prediction by under its macro-performance and periodicity [2].”
Recent Applications of Landsat 8/OLI and Sentinel-2/MSI for Land Use and Land Cover Mapping: A Systematic Review (2020, background): “[104] obtained relevant results using EVI SITS to classify summer crops in California using deep learning algorithms.”
Temporal Convolutional Neural Network for the Classification of Satellite Image Time Series (2018, background): “Consequently, these TempCNN architectures, that make the most of the temporal structure of SITS starts to be explored in remote sensing with 1D-CNNs where convolutions are applied across the temporal dimension alone [39,40], and also 3D-CNNs where convolutions are applied in both temporal and spatia”
Self-Supervised Pretraining of Transformers for Satellite Image Time Series Classification (2020, background): “specific task without the need for manual feature engineering [1, 19].”
Spotting Insects from Satellites: Modeling the Presence of Culicoides Imicola Through Deep CNNs (2019, methodology): “In such a case, the Enhanced Vegetation Index (EVI) represents a valid alternative: [12] used the EVI time-series as input for a deep learning architecture, aiming to classify summer crops (i.”
Self-Attention for Raw Optical Satellite Time Series Classification (2019, background): “, calculation of vegetation indices [4, 5, 6, 7, 8] with temporal statistics [4, 7, 9], and classification mechanisms, e.”