A method for predicting large-area missing observations in Landsat time series using spectral-temporal metrics.
Zhipeng TangHari AdhikariPetri K. E. PellikkaJanne HeiskanenPublished in: Int. J. Appl. Earth Obs. Geoinformation (2021)
Keyphrases
- detection method
- cost function
- significant improvement
- dynamic programming
- high accuracy
- high precision
- computational cost
- multispectral images
- neural network
- support vector machine svm
- feature set
- spatio temporal
- preprocessing
- image data
- unsupervised learning
- spatial information
- multispectral
- objective function
- missing values
- temporal constraints
- image segmentation