pliman_oidio.Rmd

"https://tiagoolivoto.github.io/paper_pliman/code.html#batch-processing"

```{r}
# install.packages("BiocManager")
# BiocManager::install("EBImage")
# devtools::install_github("TiagoOlivoto/pliman")
pacman::p_load(tidyverse, pliman)
```

1. Seleccione el working directory 

Ctrl + shift + h #EN MI CASO: "~/oidio_colza"

2. Importar una imagen de prueba

```{r}
img_test <- image_import("img_orig/img_04.png") 
# en este caso elijo de una subcarpeta del WD que contiene las imagenes originales
```

3. Indicar background (back), sano (healthy) y sintomatico (symptom) 

```{r}
x11() # esto hay que hacer en las ultimas versiones de RStudio por un bug que trae
healthy <- pick_palette(img_test)
symptom <- pick_palette(img_test)
back <- pick_palette(img_test)
image_combine(img_test, healthy, symptom, back, ncol = 4)
```

4. Medimos % severidad

```{r}
res <- measure_disease(
  img = img_test,
  img_healthy = healthy,
  img_symptoms = symptom,
  img_background = back,
  
  col_background = "white", # default
  col_lesions = "red", # default
  col_leaf = "green", 
  
  show_contour = FALSE, # hide the contour line
  show_original = FALSE, # create a mask
  save_image = TRUE,
  
  prefix = "proc_mask_",
  dir_processed = "img_proc"
)
```

5. Resultados

```{r}
res$severity

# comparo original con procesada
proc_test <- image_import("img_proc/proc_mask_img.jpg")

image_combine(img_test, proc_test, ncol = 2)
```

6. Batch processing 

```{r}
batch_res <- measure_disease(
  dir_original = "img_orig" ,
  pattern = "img",
  
  dir_processed = "img_proc",
  prefix = "proc_",
  
  img_healthy = healthy,
  img_symptoms = symptom,
  img_background = back,
  
  col_background = "white", # default
  col_lesions = "red", # default
  col_leaf = "green", 
  
  show_contour = FALSE, # hide the contour line
  show_original = FALSE, # create a mask
  save_image = TRUE,
  
  verbose = FALSE, 
  parallel = TRUE
)
```

Resultado como tibble 

```{r}
batch_res$severity %>% tibble
```

Extra: SAD con x niveles de las imagenes procesadas

```{r}
sad(batch_res, 4, ncol = 4)
```