Abstract
Recently, agricultural remote sensing community has endeavored to utilize the power of artificial intelligence (AI). One important topic is using AI to make the mapping of crops more accurate, automatic, and rapid. This article proposed a classification workflow using deep neural network (DNN) to produce high-quality in-season crop maps from Landsat imageries for North Dakota. We use historical crop maps from the agricultural department and North Dakota ground measurements as training datasets. Processing workflows are created to automate the tedious preprocessing, training, testing, and postprocessing workflows. We tested this hybrid solution on new images and received accurate results on major crops such as corn, soybean, barley, spring wheat, dry beans, sugar beets, and alfalfa. The pixelwise overall accuracy in all three test regions is over 82% for all land types (including noncrop land), which is the same level of accuracy as the U.S. Department of Agriculture Cropland Data Layer. The texture of DNN maps is more consistent with fewer noises, which is more comfortable to read. We find DNN is better on recognizing big farmlands than recognizing the scattered wetlands and suburban regions in North Dakota. The model trained on multiple scenes of multiple years and months yields higher accuracy than any of the models trained only on a single scene, a single month, or a single year. These results reflect that DNN can produce reliable in-season maps for major crops in North Dakota big farms and could provide a relatively accurate reference for the minor crops in scattered wetland fields.
Highlights
A GRICULTURE has been producing food, fibers, and energy to support the entire human society for thousands of years
Landsat 8 imagery is used as model inputs and Cropland Data Layer (CDL), National Agricultural Statistics Service (NASS) reports, regional crop maps and groundcollected datasets are used as training labels
We extract the pixels from CDL and make them go through a series of filtering processes to remove the ambiguous pixels which have a high chance to be incorrect
Summary
A GRICULTURE has been producing food, fibers, and energy to support the entire human society for thousands of years. The spatiotemporal and spectral variety in satellite images across the growing season should be able to help accurately recognize different types of crops (see Fig. 1). Comparing to conventional task-wise algorithms such as K-nearest neighbor, random forest (RF), and support vector machines, a well-trained DNN model seems to be able to extract more useful and common features and is expected to significantly reduce costs by being directly reused across multiple tasks. To better fit in crop mapping tasks, the SegNet model is tuned to be sensitive to minor spectral differences while maintaining accurate extraction of high-dimension large-scale contextual features. The DNN workflow runs regularly on near-real-time observed satellite images and quickly produces in-season crop maps for new Landsat scenes that are observed. This remainder of the article is organized as follows.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
