chg_traj: Calculate change-trajectory image

In azvoleff/teamlucc: TEAM land use and cover change data processing toolkit

Description

This function will calculate trajectories of land cover change using the Change Vector Analysis in Posterior Probability Space (CVAPS) approach of comparing posterior probabilities of class membership with an automatically determined threshold. Areas of no change are coded as -1. A lookup table for the codes output by chg_traj can be calculated with traj_lut.

Usage

chg_traj(chg_mag, chg_dir, chg_threshold, filename, overwrite = FALSE, ...)

Arguments

chg_mag	change magnitude RasterLayer from CVAPS
chg_dir	change direction RasterLayer from CVAPS
chg_threshold	the threshold to use as a minimum when determining change areas (can use DFPS to determine this value).
filename	filename to save the output RasterLayer to disk (optional)
overwrite	whether to overwrite existing files (otherwise an error will be raised)
...	additional parameters to pass to rasterEngine

Details

This function will run in parallel if a parallel backend is registered with foreach.

Value

a RasterLayer of change trajectories, with change trajectories coded as in the lut output by traj_lut

References

Chen, J., P. Gong, C. He, R. Pu, and P. Shi. 2003. Land-use/land-cover change detection using improved change-vector analysis. Photogrammetric Engineering and Remote Sensing 69:369-380.

Chen, J., X. Chen, X. Cui, and J. Chen. 2011. Change vector analysis in posterior probability space: a new method for land cover change detection. IEEE Geoscience and Remote Sensing Letters 8:317-321.

Examples

## Not run: 
t0_train_data <- get_pixels(L5TSR_1986, L5TSR_1986_2001_training, "class_1986",training=.6)
t0_model <- train_classifier(t0_train_data)
t0_preds <- classify(L5TSR_1986, t0_model)
t1_train_data <- get_pixels(L5TSR_2001, L5TSR_1986_2001_training, "class_2001", training=.6)
t1_model <- train_classifier(t1_train_data)
t1_preds <- classify(L5TSR_2001, t1_model)
t0_t1_chgmag <- chg_mag(t0_preds$probs, t1_preds$probs)
t0_t1_chgdir <- chg_dir(t0_preds$probs, t1_preds$probs)

lut <- traj_lut(t0_preds$codes$code, t0_preds$codes$class)
t0_t1_chgtraj <- chg_traj(lut, t0_t1_chgmag, t0_t1_chgdir, .5)

# Change areas are coded following the above lookup-table (lut):
plot(t0_t1_chgtraj)

# No change areas are -1:
plot(t0_t1_chgtraj == -1)

## End(Not run)

azvoleff/teamlucc documentation built on May 11, 2019, 5:19 p.m.