Detection of Cropland Change Using Multi-Harmonic Based Phenological Trajectory Similarity

Abstract

Accurate information on cropland changes is critical for food production and security, sustainable cropland management, and global change studies. The common change detection methods bi-temporal based, using remotely sensed imagery easily generate pseudo changes due to phenological or seasonal differences. Cropland exhibits a distinctive phenological trajectory that has strong periodic characteristics and seasonal paths. This paper proposes the use of phenological trajectory similarity to search for the overall changes between two time-series images instead of single change events between two dates of imagery. Due to the complex spectral-temporal characteristic of cropland, a phenological trajectory was constructed using a multi-harmonic model for capturing intra-annual variations. Then, phenological trajectory similarity was measured using coefficient vector difference (CVD), and used for detecting change/no-change areas when considering both the amplitude and phase difference. Finally, instead of the traditional classification method based on original images, we used the coefficient ratio vector (CRV) as the input for change type discrimination. The distance between the coefficient ratio vector (CRV) of the change pixel and of the reference change type was calculated to identify the exactly changed types. The performance of this proposed approach was tested using two sets of Landsat time-series images from 2010 and 2015. Moreover, the change area detection results of three other methods, namely, the continuous change detection and classification (CCDC), change vector analysis (CVA), and post-classification comparison (PCC), were also calculated for comparison and analysis. The results indicated that the proposed approach acquired the highest accuracy with an overall accuracy of 98.58% and a kappa coefficient of 0.82, which demonstrated that the method provides the capacity to detect real changes and estimate pseudo changes caused by season differences.

Multi-Type Forest Change Detection Using BFAST and Monthly Landsat Time Series for Monitoring Spatiotemporal Dynamics of Forests in Subtropical Wetland

“…Therefore, the phenological trajectory of all kinds of land cover categories, including farmland with complex intra-annual variations (such as double-or even triple-crop patterns within a year), must be well described for land cover change detection and classification before and after abrupt change. However, the CCDC just adopted a simple sinusoidal model for the phenological trajectory description for a variety of land cover categories, which means it cannot answer any one of four questions in Dongting Lake region with complex change processes [21,33]. Although the BFAST algorithm exhibits difficulty in interpreting the land cover categories before and after the detected abrupt changes, a multi-harmonic model adopted in BFAST could be satisfactory for the complex shape of phenological trajectory and could be applied in the Dongting Lake region.…”

A New Framework for Modelling and Monitoring the Conversion of Cultivated Land to Built-up Land Based on a Hierarchical Hidden Semi-Markov Model Using Satellite Image Time Series

“…CCDC performs the worst in comparison to the other algorithms, with the lowest overall accuracy of 92.97%, and Kappa coefficient of 0.842. This is because the simple harmonic model adopted by CCDC is incompatible with the complex phenological dynamics of cultivated land with more than one growing season [19]. The omission errors of the proposed algorithm are mainly due to the mixture of different land covers within the same pixel.…”

“…In the view of this, making use of the seasonal information contained in dense time stacks of satellite images is more advantageous [18]. Satellite image time series (SITS), which provide more vegetation phenology information, have been used for farmland change detection [19].…”

“…The basic idea of CCDC is to fit a simple harmonic model with cloudless time series in the initial phase, and then continuously detect changes when the difference between the observed and predicted values of a pixel exceeds a given threshold for consecutive times. Chen et al modified the CCDC algorithm by using a multi-harmonic model to fit the complex phenological profiles in cultivated land [19]. However, the existing studies tended to assume that the spectral-temporal characteristics of cultivated land are similar.…”

Detecting Patterns of Vegetation Gradual Changes (2001–2017) in Shiyang River Basin, Based on a Novel Framework

“…Trajectory-based change detection is an inevitable choice to do this work by constructing a 'curve' or profile of full temporal records for each pixel. Nowadays, the method has been successfully applied in forest-related change analyses [47,48], land use classification [49,50], and cropland variations [51].…”