Description Usage Arguments Value Author(s)
Takes a directory of images and directories of masks for three different classes, and then trains an SVM three-class model on that data.
1 2 3 | wrapper_train_3class_9feat(img_dir, plant_mask_dir, soil_mask_dir,
other_mask_dir, kernel = "radial", pixels_per_class = 1e+06,
probability = TRUE)
|
img_dir |
Directory where images are located |
plant_mask_dir |
Directory where masks indicating plant location are located. This directory must contain one file per image on the img_dir file with exactly the same name, and the same dimensions, but wiht only one chanel. Any non-zero value is interpreted as presence of a plant |
soil_mask_dir |
Directory where masks indicating soil location are located. This directory must contain one file per image on the img_dir file with exactly the same name, and the same dimensions, but wiht only one chanel. Any non-zero value is interpreted as presence of a soil |
other_mask_dir |
Directory where masks indicating other location are located. This directory must contain one file per image on the img_dir file with exactly the same name, and the same dimensions, but wiht only one chanel. Any non-zero value is interpreted as presence of a other. |
kernel |
Type of kernel to use for SVM. See documentation from e1071 |
pixels_per_class |
Not all pixels need to be used. Guud accuracy can be achieved with few pixels per class. Specify the number here. |
probability |
Logical indicating whether the SVM model must include probabilities or just classification |
A list with elements m1 which contains the SVM trained model that can be used for prediction, and Dat element which is a data.frame containing the Data used to train the model
Sur Herrera Paredes
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.