Takes a directory of images, and tries to predict on all images on that directory
wrapper_predictdir_9feat(img_dir, overlaydir, maskdir, m1, size_threshold = 0)
| img_dir | Directory name where images files are  | 
    
|---|---|
| overlaydir | directory anem where to save the overlay images  | 
    
| maskdir | directory name where the mask files are saved  | 
    
| m1 | trained SVM model from e1071 used for prediction  | 
    
| size_threshold | the minimum size of of a contiguous set of selected threshold for that group to be included. Size is in number of pixels.  | 
    
Data.frame showing file names processed and object size in pixels from each file