Corn Phenology Detection Using the Derivative Dynamic Time Warping Method and Sentinel-2 Time Series

Abstract

Accurate determination of crop phenology information is essential for effective field management and decision-making processes. Remote sensing time series analyses are widely employed to extract the phenological stages. Each crop’s phenological stage has its unique characteristic on the crop plant, while the satellite-derived crop phenology refers to some key transition dates in time series satellite observations. Current techniques primarily estimate specific phenological stages by detecting points with distinctive features on the remote sensing time series curve. But these stages may be different from the Biologische Bundesanstalt, Bundessortenamt and CHemical Industry (BBCH) scale, which is commonly used to identify the phenological development stages of crops. Moreover, when aiming to extract various phenological stages concurrently, it becomes necessary to adjust the extraction strategy for each unique feature. This need for distinct strategies at each stage heightens the complexity of simultaneous extraction. In this study, we utilize the Sentinel-2 Normalized Difference Vegetation Index (NDVI) time series data and propose a phenology extraction framework based on the Derivative Dynamic Time Warping (DDTW) algorithm. This method is capable of simultaneously extracting complete phenological stages, and the results demonstrate that the Root Mean Square Errors (RMSEs, days) of detected phenology on the BBCH scale for corn were less than 6 days overall.

C-DTW for Human Action Recognition Based on Nanogenerator

“…DDTW (d) and shapeDTW (f) achieve similar results to C-DTW, as both methods incorporate shape information. DDTW transforms the original time series into high-level features containing shape information by using the differences between the sequences [ 40 ], while shapeDTW considers the shape correspondence of each sub-sequence using a sliding window [ 29 ]. On the other hand, WDTW (c) and softDTW (e) do not take shape features into account.…”

“…The reason for this lies in the inherent limitations of DTW, which stem from the features it considers. DTW only takes into account the y-axis values of datapoints, and does not effectively handle their shapes [ 40 ]. Thus, even with the selection of DTW alignment it is not possible to achieve better results on the basis of alignments that inherently do not take shape relationships into account.…”