Add files via upload

multi_vi.R

@@ -0,0 +1,131 @@
+rm(list = ls())
+
+library(tidyverse)
+library(rio)
+library(raster)
+library(EBImage)
+
+
+# Image Import Path
+floder_path <- "Rdata/"
+
+# Chose Multispectral Sensor 
+sensor_type <- "RedEdge" #"DJI" or "PARROT" or "RedEdge" or "Sentinel-2"
+
+# Defind DN Value or Reflectance
+raster_type <- "DN" #"RE" or "DN"
+
+# Result Save Path
+output_path <- "Rresult/"
+
+
+
+
+
+res <- data.frame()
+
+# Defind Each Bands(num or NA)
+if (sensor_type == "RedEdge") {
+  GRE <- 1
+  NIR <- 4
+  BLU <- NA
+  RED <- 2
+  REG <- 3
+} else if (sensor_type == "DJI"){
+  GRE <- 1
+  NIR <- 4
+  BLU <- NA
+  RED <- 2
+  REG <- 3
+} else if (sensor_type == "RedEdge"){
+  GRE <- 1
+  NIR <- 4
+  BLU <- NA
+  RED <- 2
+  REG <- 3
+} else {
+  GRE <- 3
+  NIR <- 8
+  BLU <- 2
+  RED <- 4
+  REG <- 6 #740nm
+}
+
+
+file_path <- list.files(floder_path, full.names = T)
+file_name <- list.files(floder_path)
+dir_name <- floder_path
+
+out_floder <- paste0(output_path, substr(floder_path, 7, nchar(floder_path)))
+dir.create(out_floder)
+
+
+res <- data.frame()
+
+for (i in 1:length(file_path)) {
+  # i <- 2
+  img_file <- file_path[i]
+  name_file <- file_name[i]
+  img <- brick(img_file)
+  plot(img)
+
+  pixels <- dim(img)[1]*dim(img)[2]
+
+  img_name <- substr(name_file, 1, nchar(name_file)-4)
+
+  img_out <- paste0(out_floder, img_name)
+  dir.create(img_out)
+
+  # Extract Each Bands
+  if (raster_type == "DN") {
+    rv = 32768
+  } else {
+    rv = 1
+  }
+
+  calculate_band <- function(index) {
+    if (!is.na(index)) {
+      return(img[[index]] / rv)
+    } else {
+      return(NA)
+    }
+  }
+
+  gre <- calculate_band(GRE)
+  nir <- calculate_band(NIR)
+  blu <- calculate_band(BLU)
+  red <- calculate_band(RED)
+  reg <- calculate_band(REG)
+
+
+  vi_fun <- import("vi_fun_multi.csv")
+  vi_list <- vi_fun[, 1]
+
+  df <- data.frame(plot = img_name)
+  for (i in 1:length(vi_list)) {
+    vi_name <- as.character(vi_list[i])
+    fun <- filter(vi_fun, index == vi_name)
+    # Caculate vegetation index image
+    vi = eval(parse(text = fun[, 2]))
+    # Output vegetation index image
+    # vi_img <- as.Image(vi)
+    plot(vi)
+    writeRaster(vi, paste0(img_out, "/", vi_name,".tif"), overwrite = TRUE)
+    # Caculate vegetation indeies
+    vi_array <- sort(as.numeric(vi@data@values))
+    vi_v <- mean(vi_array[round(pixels*0.1):round(pixels*0.9)])
+    vi_df <- data.frame(vi_v)
+    colnames(vi_df) <- vi_name
+
+    df <- cbind(df, vi_df)
+  }
+
+  res <- rbind(res, df)
+}
+
+
+
+
+
+write.csv(res, paste0(out_floder, "vi.csv"), row.names = F)
+