Here is an aticle to guide you step by step through the extraction of crowns scale data. All you need to start it is a RGB orthomosaics time serie and a shapefile of the crowns delineations. To run the following code you have to download the test data (link is coming).
############## Load libraries --------------------------------------------------
library(canObsR)
library(tidyverse)
library(sf)
library(openxlsx)Set parameters
Set the path to the CanObsR_data folder, the site name, the Tmpfiles_directory if you want to store the temporary files in a specific folder and N_cores the number of cores you want to allocate for the parallelized functions*.
Note* : The functions extract_crownsImages() and
extract_rgbValues() can produce a lot of temporary files
and need several GB of memory depending on the size, the resolution and
the number of mosaics. You can specify where you want to store the
temporary files.
############## Set parameters --------------------------------------------------
# Path to the main directory
path = 'XXXX/canObsR_data'
# Site
site = 'Bouamir'
# Path to store the temp files
Tmpfiles_directory = 'XXX/R_tmpdir' # NULL if you want to use the default folder
# Number of cores used for the paralelized functions
N_cores = 10If you are using the test data just run the following code to get the path to the RGB mosaics (imgs) and the dates of the images (dates).
# Path to the mosaics
imgs <- list.files(file.path(path,'3_orthomosaics_aligned'), full.names = T)
# Extract the dates of the images
dates <-
extr_dates(names_img = basename(imgs), n = 2, sep = '', extension = '.tif')First step : Test the crowns shapefile compatible
The first step is to check the crowns file compatibility with the
function check_crownsFile. To be compatible with the
package functions the shapefile must have :
- at least the variables ‘id’, ‘species’, ‘genus’ and ‘family’
- ‘species’, ‘genus’ and ‘family’ should not have NA. Replace them by ‘indet’
- ‘id’ should not have duplicated values.
############## STEP 1 Check and modify the crowns file ------------------------
# Check the crowns file compatibility
check_crownsFile(
path_crowns = file.path(path,'crowns.gpkg')
) # Output : Image 1 (1)
#------------------------------------------------------------------------------#
#------------------------------------------------------------------------------#
# Rename variables
sf::read_sf(file.path(path,'crowns.gpkg')) %>% # Load the file
rename(
family = tax_fam, # Rename the corresponding variable as 'family'
genus = tax_gen, # Rename the corresponding variable as 'genus'
species = tx_sp_lvl # Rename the corresponding variable as 'species'
) %>%
sf::write_sf(.,file.path(path,'crowns_corrected.gpkg')) # And save the file
# Check the new crowns file compatibility
check_crownsFile(
path_crowns = file.path(path,'crowns_corrected.gpkg')
) # Output : Image 1 (2)
#------------------------------------------------------------------------------#
#------------------------------------------------------------------------------#
# Rename NA to 'indet'
sf::read_sf(file.path(path,'crowns_corrected.gpkg')) %>% # Load the new file
mutate(
# Rename NA to 'indet' in the family variable
family = case_when(is.na(family)~ 'indet', TRUE ~ family),
# Same for genus
genus = case_when(is.na(genus)~ 'indet', TRUE ~ genus),
# Same for species
species = case_when(is.na(species)~ paste(genus, 'indet'), TRUE ~ species)
) %>%
sf::write_sf(.,file.path(path,'crowns_corrected.gpkg')) # And save the file
# Re-check the new crowns file compatibility
check_crownsFile(
path_crowns = file.path(path,'crowns_corrected.gpkg')
) # Output : Image 1 (3)
check_crownsFile outputs
Second step : Extract crowns images per date
The second step is to extract individual crown images per date with
the function extract_crownsImages. The function can be used
in parallelization when N_cores is superior at 1 and can use a lot of
memory (several GB). It is better to specify where to store the
temporary files to be sur to have enought space. If you use the test
data, it can take about 15 minutes (or more).
############## STEP 2 Extract crowns images ------------------------------------
extract_crownsImages(
path_images = imgs, # Path to mosaics
path_crowns = file.path(path,'crowns_corrected.gpkg'), # Path to crowns file
out_dir_path = file.path(path,'4_crowns_img'), # Path to output folder
tempdir_custom = Tmpfiles_directory, # Path to tmp file directory
site = NULL, # If NULL, will be extract from the path_images
dates = NULL, # If NULL, will be extract from the path_images
N_cores = N_cores, # Number of cores
width = 720, # By default it is 720
height = 825 # By default it is 720
)
extract_crownsImages outputs files
Third step : Extract crowns images per date
############## STEP 3 Extract spectral indices from mosaics at the crown scale -
rgb_data <- extract_rgbValues (
path_images = imgs, # Path to mosaics
path_crowns = file.path(path,'crowns_corrected.gpkg'), # Path to crowns file
out_dir_path = path, # Path to output folder
tempdir_custom = Tmpfiles_directory, # Path to tmp file directory
ncor = N_cores, # Number of cores
sites = NULL, # If NULL, will be extract from the path_images
dates = NULL, # If NULL, will be extract from the path_images
file_type = '.xlsx' # Could be '.csv' or '.xlsx' or 'RData'
)
extract_rgbValues outputs files
Fouth step : Create the labeling file and label images
create_labelingFile will create a xlsx file that can
directly be used by a shiny application (shiny_labels) to
do the labeling.
############## STEP 4 Create file to do the labeling --------------------------
create_labelingFile(
path_crowns = file.path(path,'crowns_corrected.gpkg'), # Path to crowns file
site = site, # Site name
dates = dates, # Dates of the RGB survey
out_dir_path = path) # Path to output folder
# open it
openXL(file.path(path,paste(site,'labelingFile',paste0(format(as.Date(Sys.Date(),format="%Y-%m-%d"), format = "%Y%m%d"), '.xlsx'), sep = '_' )))
############## STEP 5 Fill the labelingFile with shiny app --------------------
# Import the labelingFile
labeling_file <- file.path(
path,
paste('Bouamir_labelingFile',
paste0(
format(as.Date(Sys.Date(),format="%Y-%m-%d"), format ="%Y%m%d"),
'.xlsx'),
sep = '_' )
)
# Start the app
shiny_labels(
data_labeling = labeling_file,
newFile = file.path(path,paste(site,'labelingFile_update.xlsx')),
imgFolder = file.path(path,'4_crowns_img') # Path to the crowns images folder
) 
shiny_visualisation