Hyperspectral data processing
What are hyperspectral images?
Hyperspectral sensors provide optical radiance data in >200 channels from 350-2500 nm with a spectral resolution of < 10 nm, generating images of contiguous spectral channels often spanning 400–1050 nm (visible to near-infrared) or 400–2500 nm (visible to shortwave-infrared) wavelength range (Figure 1), that can be use to characterize the chemistry of the vegetation, and allow a better discrimination between subtle physiological differences among plant species.
What types of variables can be extracted from hyperspectral data?
Hyperspectral data allows to trace, via spectral separability, subtle differences in the leaf pigment, nutrient and structural properties of the vegetation at different levels of aggregation. Such group-specific spectral properties are known as reflectance signatures that can be use to map and monitor vegetation changes from above.
Hyperspectral data has multiple applications, from allowing the differentiation between bare ground and ground with vegetation cover, and invasive vs. native species, to the accurate identification and mapping of single invasive and native species.
The dataset
To create the dataset provided in this section of the tutorial, we generated an artificial hyperspectral image (Figure 2) from the original APEX data of Sylt island (Germany) collected and preprocessed by the Flemish Institute for Technological Research VITO. The artificial image was created by taking a buffer area from the center of the calibration, validation and background points, and then rearranging the buffer squares together with the corresponding points, which resulted in an image with the real hyperspectral values (grid cells) at each point but reorganize in a smaller file (Figure 3). You can download the dataset for this tutorial from HERE.
Setting up GRASS GIS within R
It is recommended to first create a folder where the hyperspectral image is located and the outputs and results are saved. Now you can set the working directory in R using the folder you just created:
setwd("/home/garzonc/DIARS/")
Load the rgrass7 library and create a directory that will contain the GRASS GISDBASE for the working session:
library(rgrass7)
dir.create("grassdata")
To initiate GRASS into your R session you need to set the gisBase argument to the GRASS binaries and libraries directory path, this should be something like C:\Program Files\GRASS GIS 7.00beta4 if your are using windows (click HERE for help on how to setup rgrass7 in other operating systems). Note that this will create a new GRASS location called Sylt_island inside the GRASS GISDBASE.
myGRASS <- "/home/garzonc/src/grass7_trunk/dist.x86_64-pc-linux-gnu"
myGISDbase <- "/home/garzonc/Documents/grassData" #Set the directory path to your GISDbase
myLocation <- "DIARS" #Name the Location
myMapset <- "PERMANENT" #Name the mapset
initGRASS(myGRASS, home = tempdir(), gisDbase = myGISDbase, location = myLocation,
mapset = myMapset, override = TRUE)## gisdbase /home/garzonc/Documents/grassData
## location DIARS
## mapset PERMANENT
## rows 170
## columns 185
## north 6080249
## south 6079943
## west 463975.2
## east 464308.2
## nsres 1.8
## ewres 1.8
## projection +proj=utm +no_defs +zone=32 +a=6378137 +rf=298.257223563
## +towgs84=0.000,0.000,0.000 +to_meter=1execGRASS("g.region", parameters = list(n = "6080248.8", s = "6079942.8", e = "464308.2",
w = "463975.2", res = "1.8"))
Importing the hyperspectral data into R
The image provided is ready to use, so you can import it directly into R. Remember to have the image stored in a new folder you created at the beginning of this tutorial. To manage this image (raster file) in R is necessary to install and load the rasterpackage:
library(raster)
Plot the one of the bands (166) of the hyperspectral image:
setwd("/home/garzonc/Desktop/DIARS/")
filelist <- dir(pattern = ".tiff")
mo <- raster(filelist[5], band = 150, package = "raster")
plot(mo, col = colorRampPalette(c("blue", "green", "red"))(255), legend = FALSE)
Exploring the hyperspectral data
Hyperspectral images have many advantages due to the large amount of information contained in each pixel (in this case 248 bands). Lets visualize how all these data might look like. To do that first lets install the raster, rgdal and foreach:
library(raster)
library(rgdal)
library(foreach)
Now lets extract a sample of the bands using the point dataset provided. First, import the point dataset. Lets plot the curve of spectral responses for a sample calibration and background points, that is, the percent reflectance values per point across the range of wavelengths in the hyperspectral image.
setwd("/home/garzonc/Desktop/DIARS/")
spec <- read.table("PointBackground.csv", header = T, sep = " ")[, -c(2:3)]
library(reshape2)
library(ggplot2)
sample <- melt(spec, id = c("type"))
p <- ggplot(sample, aes(sample[, 2], sample[, 3])) + geom_boxplot(aes(fill = sample[,
2]), outlier.colour = NULL) + theme(axis.text.x = element_text(angle = 90,
hjust = 0), axis.title = element_text(face = "bold", size = 14), title = element_text(face = "bold",
size = 16), legend.position = "none") + labs(title = paste("Spectral response"),
x = "Wavelength, nm", y = "Response")
p
As you can see from the graph there is a large amount of data/information, this provides a huge potential and at the same time posits major limitations in time and processing power. One of the techniques most commonly used to reduce the dimension of the data, without loosing relevant information, is the Principal Components Analysis (PCA). Some of the variables (bands) might be highly correlated among them and so PCA allows extracting the most relevant variables (in form of components) from a large set of variables available in a data set. Calculate the PCA using the prcomp function using the already extracted values from the previous step:
spec1 <- as.matrix(spec[, -c(1)])
pca <- prcomp(spec1, scale. = T, center = T)
summary(pca) # Identify the coverage of variance in the dataset by individual principal components ## Importance of components:
## PC1 PC2 PC3 PC4 PC5 PC6
## Standard deviation 14.6289 5.4589 1.64638 0.79349 0.51266 0.44784
## Proportion of Variance 0.8629 0.1202 0.01093 0.00254 0.00106 0.00081
## Cumulative Proportion 0.8629 0.9831 0.99402 0.99655 0.99761 0.99842
## PC7 PC8 PC9 PC10 PC11 PC12
## Standard deviation 0.30434 0.26205 0.20100 0.19221 0.17002 0.14674
## Proportion of Variance 0.00037 0.00028 0.00016 0.00015 0.00012 0.00009
## Cumulative Proportion 0.99880 0.99907 0.99924 0.99939 0.99950 0.99959
## PC13 PC14 PC15 PC16 PC17 PC18
## Standard deviation 0.14277 0.11129 0.10043 0.09220 0.07906 0.07331
## Proportion of Variance 0.00008 0.00005 0.00004 0.00003 0.00003 0.00002
## Cumulative Proportion 0.99967 0.99972 0.99976 0.99980 0.99982 0.99984
## PC19 PC20 PC21 PC22 PC23 PC24
## Standard deviation 0.06997 0.06672 0.06073 0.05182 0.04905 0.04751
## Proportion of Variance 0.00002 0.00002 0.00001 0.00001 0.00001 0.00001
## Cumulative Proportion 0.99986 0.99988 0.99990 0.99991 0.99992 0.99992
## PC25 PC26 PC27 PC28 PC29 PC30
## Standard deviation 0.04510 0.04240 0.04024 0.03719 0.03292 0.0314
## Proportion of Variance 0.00001 0.00001 0.00001 0.00001 0.00000 0.0000
## Cumulative Proportion 0.99993 0.99994 0.99995 0.99995 0.99996 1.0000
## PC31 PC32 PC33 PC34 PC35 PC36
## Standard deviation 0.02865 0.02789 0.02577 0.02367 0.02237 0.02044
## Proportion of Variance 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
## Cumulative Proportion 0.99996 0.99997 0.99997 0.99997 0.99997 0.99998
## PC37 PC38 PC39 PC40 PC41 PC42
## Standard deviation 0.01967 0.01863 0.01793 0.01627 0.01559 0.01436
## Proportion of Variance 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
## Cumulative Proportion 0.99998 0.99998 0.99998 0.99998 0.99998 0.99998
## PC43 PC44 PC45 PC46 PC47 PC48
## Standard deviation 0.0142 0.01386 0.01324 0.01294 0.01261 0.01228
## Proportion of Variance 0.0000 0.00000 0.00000 0.00000 0.00000 0.00000
## Cumulative Proportion 1.0000 0.99998 0.99999 0.99999 0.99999 0.99999
## PC49 PC50 PC51 PC52 PC53 PC54
## Standard deviation 0.01152 0.0114 0.01083 0.01052 0.01004 0.009601
## Proportion of Variance 0.00000 0.0000 0.00000 0.00000 0.00000 0.000000
## Cumulative Proportion 0.99999 1.0000 0.99999 0.99999 0.99999 0.999990
## PC55 PC56 PC57 PC58 PC59
## Standard deviation 0.009523 0.009263 0.009027 0.008808 0.008628
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 0.999990 0.999990 0.999990 0.999990 0.999990
## PC60 PC61 PC62 PC63 PC64
## Standard deviation 0.008437 0.008301 0.008261 0.007928 0.007763
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 0.999990 0.999990 0.999990 0.999990 0.999990
## PC65 PC66 PC67 PC68 PC69
## Standard deviation 0.007553 0.007508 0.007444 0.007352 0.007197
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 0.999990 0.999990 0.999990 0.999990 0.999990
## PC70 PC71 PC72 PC73 PC74
## Standard deviation 0.007123 0.006974 0.006858 0.006759 0.006704
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 0.999990 0.999990 0.999990 0.999990 0.999990
## PC75 PC76 PC77 PC78 PC79
## Standard deviation 0.006593 0.006395 0.006327 0.006293 0.00621
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.00000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.00000
## PC80 PC81 PC82 PC83 PC84
## Standard deviation 0.006201 0.006057 0.005945 0.005805 0.005748
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC85 PC86 PC87 PC88 PC89
## Standard deviation 0.005657 0.005512 0.005457 0.005419 0.005338
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC90 PC91 PC92 PC93 PC94
## Standard deviation 0.005291 0.005137 0.005064 0.005044 0.004964
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC95 PC96 PC97 PC98 PC99
## Standard deviation 0.004918 0.004842 0.004777 0.004663 0.004617
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC100 PC101 PC102 PC103 PC104
## Standard deviation 0.004545 0.004541 0.004475 0.004391 0.004334
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC105 PC106 PC107 PC108 PC109
## Standard deviation 0.004276 0.00422 0.004009 0.003978 0.003905
## Proportion of Variance 0.000000 0.00000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.00000 1.000000 1.000000 1.000000
## PC110 PC111 PC112 PC113 PC114
## Standard deviation 0.003859 0.003791 0.003772 0.003713 0.003611
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC115 PC116 PC117 PC118 PC119
## Standard deviation 0.003506 0.003463 0.003408 0.003348 0.003327
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC120 PC121 PC122 PC123 PC124
## Standard deviation 0.003295 0.003176 0.003065 0.003008 0.002977
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC125 PC126 PC127 PC128 PC129
## Standard deviation 0.002946 0.002892 0.002821 0.00269 0.002632
## Proportion of Variance 0.000000 0.000000 0.000000 0.00000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.00000 1.000000
## PC130 PC131 PC132 PC133 PC134
## Standard deviation 0.002616 0.00257 0.002527 0.002444 0.002424
## Proportion of Variance 0.000000 0.00000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.00000 1.000000 1.000000 1.000000
## PC135 PC136 PC137 PC138 PC139 PC140
## Standard deviation 0.002346 0.002226 0.002198 0.002107 0.00202 0.00197
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.00000 0.00000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.00000 1.00000
## PC141 PC142 PC143 PC144 PC145
## Standard deviation 0.001947 0.001842 0.001793 0.001749 0.001617
## Proportion of Variance 0.000000 0.000000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.000000 1.000000
## PC146 PC147 PC148 PC149 PC150 PC151
## Standard deviation 0.001548 0.0015 0.00145 0.001378 0.001346 0.001187
## Proportion of Variance 0.000000 0.0000 0.00000 0.000000 0.000000 0.000000
## Cumulative Proportion 1.000000 1.0000 1.00000 1.000000 1.000000 1.000000
## PC152 PC153 PC154 PC155 PC156
## Standard deviation 0.001161 0.001089 0.001013 0.0009727 0.0009526
## Proportion of Variance 0.000000 0.000000 0.000000 0.0000000 0.0000000
## Cumulative Proportion 1.000000 1.000000 1.000000 1.0000000 1.0000000
## PC157 PC158 PC159 PC160 PC161
## Standard deviation 0.0008628 0.0007341 0.0006595 0.0006137 0.00057
## Proportion of Variance 0.0000000 0.0000000 0.0000000 0.0000000 0.00000
## Cumulative Proportion 1.0000000 1.0000000 1.0000000 1.0000000 1.00000
## PC162 PC163 PC164 PC165 PC166
## Standard deviation 0.0005528 0.0005001 0.000422 0.0004137 0.0004099
## Proportion of Variance 0.0000000 0.0000000 0.000000 0.0000000 0.0000000
## Cumulative Proportion 1.0000000 1.0000000 1.000000 1.0000000 1.0000000
## PC167 PC168 PC169 PC170 PC171
## Standard deviation 0.0003796 0.0003488 0.0002875 0.0002759 0.0002324
## Proportion of Variance 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000
## Cumulative Proportion 1.0000000 1.0000000 1.0000000 1.0000000 1.0000000
## PC172 PC173 PC174 PC175 PC176
## Standard deviation 0.0002077 0.0001912 0.0001456 0.0001319 0.0001211
## Proportion of Variance 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000
## Cumulative Proportion 1.0000000 1.0000000 1.0000000 1.0000000 1.0000000
## PC177 PC178 PC179 PC180 PC181
## Standard deviation 9.576e-05 8.971e-05 7.498e-05 5.807e-05 5.614e-05
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC182 PC183 PC184 PC185 PC186
## Standard deviation 4.066e-05 3.703e-05 3.482e-05 2.242e-05 1.638e-05
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC187 PC188 PC189 PC190 PC191
## Standard deviation 1.523e-05 1.441e-05 1.336e-05 1.311e-05 7.177e-06
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC192 PC193 PC194 PC195 PC196
## Standard deviation 6.44e-06 4.416e-06 4.092e-06 3.426e-06 2.452e-06
## Proportion of Variance 0.00e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.00e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC197 PC198 PC199 PC200 PC201
## Standard deviation 1.699e-06 1.199e-06 6.6e-07 4.673e-07 3.114e-07
## Proportion of Variance 0.000e+00 0.000e+00 0.0e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.0e+00 1.000e+00 1.000e+00
## PC202 PC203 PC204 PC205 PC206
## Standard deviation 2.385e-07 1.683e-07 1.583e-07 1.475e-07 1.433e-07
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC207 PC208 PC209 PC210 PC211
## Standard deviation 1.387e-07 1.372e-07 1.338e-07 1.311e-07 1.255e-07
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC212 PC213 PC214 PC215 PC216
## Standard deviation 1.241e-07 1.227e-07 1.204e-07 1.19e-07 1.18e-07
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.00e+00 0.00e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.00e+00 1.00e+00
## PC217 PC218 PC219 PC220 PC221
## Standard deviation 1.165e-07 1.138e-07 1.119e-07 1.097e-07 1.083e-07
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC222 PC223 PC224 PC225 PC226
## Standard deviation 1.06e-07 1.013e-07 9.363e-08 9.127e-08 7.368e-08
## Proportion of Variance 0.00e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.00e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC227 PC228 PC229 PC230 PC231
## Standard deviation 6.779e-08 6.31e-08 5.982e-08 5.89e-08 5.669e-08
## Proportion of Variance 0.000e+00 0.00e+00 0.000e+00 0.00e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.00e+00 1.000e+00 1.00e+00 1.000e+00
## PC232 PC233 PC234 PC235 PC236
## Standard deviation 5.448e-08 5.162e-08 5.034e-08 4.975e-08 4.88e-08
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.00e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.00e+00
## PC237 PC238 PC239 PC240 PC241
## Standard deviation 4.691e-08 4.627e-08 4.535e-08 4.334e-08 4.252e-08
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC242 PC243 PC244 PC245 PC246
## Standard deviation 4.088e-08 4.048e-08 3.951e-08 3.826e-08 3.728e-08
## Proportion of Variance 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00
## Cumulative Proportion 1.000e+00 1.000e+00 1.000e+00 1.000e+00 1.000e+00
## PC247 PC248
## Standard deviation 3.56e-08 3.516e-08
## Proportion of Variance 0.00e+00 0.000e+00
## Cumulative Proportion 1.00e+00 1.000e+00pca_red <- prcomp(spec1, scale. = T, center = T, tol = 0.01) # Calculate only first components, tol=0
pca_red$sdev[1:11]## [1] 14.6289380 5.4588915 1.6463772 0.7934939 0.5126644 0.4478358
## [7] 0.3043380 0.2620548 0.2009957 0.1922074 0.1700204The PCA results shows the variance contained in the first component and the weight of each variable (band) contributing to the component. Based on those results we can exclude the bands that contribute the least.
The artificial image is already provided as part of the tutorial dataset so you don’t need to perform the process explained below. This is only aimed to guide those interested in developing an artificial image with their own data
Creating an artificial hyperspectral image - Optional (advanced GRASS GIS and bash programming skills)
The artificial image developed as part of this dataset was created using GRASS GIS software. The following guide requires knowledge of GRASS GIS and bash programming
The workflow presented here (Figure 4) summarizes the steps followed to generate an artificial image:
After location and mapset have been set, import the hyperspectral image and the point data which in this case correspond to the validation, calibration and background locations previously selected.
r.in.gdal -o input=mosaic_Sylt_20140716_v3.1 output=Sylt_mosaic
v.in.ogr input=/home/user/val_pointsDataset.shp output=val_points
Create buffer areas around the points in the point file. Make sure the point is in the middle of a pixel to ensure selecting the exact same number of pixels per buffer area.
%Convert points to raster to select the pixel that overlaps with the point
v.to.rast input=ALLpoints@Sylt type=point output=ALLpoints use=cat
%Convert raster back to points that will be located exactly at the center of the pixel
r.to.vect -v input=ALLpoints@Sylt output=ALLpointsUP type=point
%Create buffer area around the of 4.5 m distance around the recentered points
v.buffer -t input=ALLpointsUP@Sylt output=ALL_buffers distance=4.5
All the buffers are saved as one single layer, but we need to have each buffer area as a separate file. You can do it one by one, using the ` or you can use the following batch script:
for i in `seq 1 646`; do
v.extract input=ALL_buffers cats=`$`{i} output=Bufdata`$`{i} --o
done
Now each buffer area is in a single file but the buffer areas are circular, to convert them to squares you can use the following bash script:
for i in `seq 1 646`; do
num=`expr $num + 1`
g.region vector=Bufdata`$`{i} res=`$`PIXELSIZE -p
v.in.region all_bufferUP_`$`num --o
done%Set region back to the entire image
g.region raster=mosaic_Sylt_20140716_v3.1 -p
Now we can create an empty grid in which the artificial image is going to be put together. The size of the cell (box=9,9) must match the size of the buffer areas and the number of grid cells (grid=34,35) must add up to the total number of points/buffers.
v.mkgrid –o map=ALL_buffers grid=34,35 position=coor coordinates=699402.6,6912889.5 box=9,9
for i in `seq 1 646`; do
v.extract input=ALL_buffers@Sylt cats=`$`{i} output=grid_`$`{i} --o
done
x=0
for a in `seq 1 248`; do
for i in `seq 1 646`; do
g.region vector=all_bufferUP_`$`{i} -ap
r.mask vect=all_bufferUP_`$`{i}
echo "mosaic_Sylt_20140716_v3.`$`{a}"
r.mapcalc expression="rast_buffer_`$`{a}_`$`{i} = mosaic_Sylt_20140716_v3.`$`{a}" --o
r.mask -r
x=`expr $x + 1`
echo "grid `$`x"
g.region vector=grid_`$`x
g.region res=1.8 -ap
eval `g.region -g`
r.region map=rast_buffer_`$`{a}_`$`{i} n=`$`n s=n-9 w=e-9 e=`$`e
g.region raster=mosaic -p
echo "bufffer `$`{i} band `$`{a}"
done
echo "Band `$`{a} and `$`x grids done"
x=0
echo "End `$`SECONDS seconds"
done
for i in `seq 1 248`; do
for a in `seq 1 248`; do
MAPS1=`g.list -e type=raster sep=, pat="^rast_buffer_${a}_([1-9]|[1-9][0-9]|[1-5][0-9][0-9]|6[0-4][0-9])$"`
g.region raster=$MAPS1
r.patch in=$MAPS1 out=grid_mosaicC_1_${a} --o
MAPS2=`g.list -e type=raster sep=, pat="^rast_buffer_${a}_([2-3][0-9][0-9])$"`
g.region raster=$MAPS2
r.patch in=$MAPS2 out=grid_mosaicC_2_${a} --o
MAPS3=`g.list -e type=raster sep=, pat="^rast_buffer_${a}_([4-5][0-9][0-9])$"`
g.region raster=$MAPS3
r.patch in=$MAPS3 out=grid_mosaicC_3_${a} --o
MAPS4=`g.list -e type=raster sep=, pat="^rast_buffer_${a}_(6[0-4][0-9])$"`
g.region raster=$MAPS4
r.patch in=$MAPS4 out=grid_mosaicC_4_${a} --o
echo "grid_mosaicC_${a}"
done
MAPS=`g.list -e type=raster sep=, pat="^grid_mosaicC_[1-4]_${i}$"`
g.region raster=$MAPS
r.patch in=$MAPS out=grid_mosaicC_all_${i} --o
echo "grid_mosaic_${i}"
doneThe resulting artificial raster bands need to be renumbered to avoid misplacing when exporting the image as a tiff file. We will add 00 in front to have it organized in the right band order using the following bash script:
prefix="mosaicC_full_"
for map in `g.list type=raster pattern=${prefix}*`; do
part=`echo $map | cut -d "_" -f 3`
numero=`echo $part | awk '{ printf("%04d\n", $1) }'`
names="${prefix}${numero}"
g.rename raster=$map,$names
doneWe can now verify that the hyperspectral artificial image has the correct band order, and if so, export it:
MAPS=g.list type=raster sep=, pat="mosaicC_full_*" i.group group=mosaic_datasetComp input=$MAPS
i.spectral -g group=mosaic_datasetComp@compiegne coordinates=699412.07161,6912899.83863
r.out.gdal in=mosaic_datasetComp out=mosaic_datasetComp.tiff -c –o
Back to the top
Authors: Carol Garzon-Lopez, Ruben Van De Kerchove, Duccio Rocchini & Jonathan Lenoir
Date: 2018-08-24
Built with RMarkdown and hosted on Github Pages.