ee.Algorithms.TemporalSegmentation.Ccdc

  • The CCDC algorithm detects temporal breakpoints in image collections using harmonic functions.

  • Fit coefficients are produced for all bands, but breakpoint detection can be band-specific.

  • The algorithm is based on the Continuous Change Detection and Classification method by Zhu and Woodcock (2014).

 Implements the Continuous Change Detection and Classification temporal breakpoint algorithm. This algorithm finds temporal breakpoints in an image collection by iteratively fitting harmonic functions to the data. Fit coefficients are produced for all input bands, but the bands used for breakpoint detection can be specified with the 'breakpointBands' argument.

For more details, see Zhu, Z. and Woodcock, C.E., 2014. Continuous change detection and classification of land cover using all available Landsat data. Remote sensing of Environment, 144, pp.152-171.

UsageReturns
ee.Algorithms.TemporalSegmentation.Ccdc(collection, breakpointBands, tmaskBands, minObservations, chiSquareProbability, minNumOfYearsScaler, dateFormat, lambda, maxIterations)Image
ArgumentTypeDetails
collectionImageCollectionCollection of images on which to run CCDC.
breakpointBandsList, default: nullThe name or index of the bands to use for change detection. If unspecified, all bands are used.
tmaskBandsList, default: nullThe name or index of the bands to use for iterative TMask cloud detection. These are typically the green band and the SWIR1 band. If unspecified, TMask is not used. If specified, 'tmaskBands' must be included in 'breakpointBands'.
minObservationsInteger, default: 6The number of observations required to flag a change.
chiSquareProbabilityFloat, default: 0.99The chi-square probability threshold for change detection in the range of [0, 1].
minNumOfYearsScalerFloat, default: 1.33Factors of minimum number of years to apply new fitting.
dateFormatInteger, default: 0The time representation to use during fitting: 0 = jDays, 1 = fractional years, 2 = unix time in milliseconds. The start, end and break times for each temporal segment will be encoded this way.
lambdaFloat, default: 20Lambda for LASSO regression fitting. If set to 0, regular OLS is used instead of LASSO.
maxIterationsInteger, default: 25000Maximum number of runs for LASSO regression convergence. If set to 0, regular OLS is used instead of LASSO.