from pictureclassifier import Picture_classifier

from objectclassifier import Object_classifier

def main():

pic_cla = None

keep = True

while keep:

print("\nGive 'T' if you want to train model, \n'S' if you want to sort pictures or \n'Q' if you want to quit:\n")

char = input()

if char == "T":

print("Write the path of the root directory, where the training material is.\nYou need to have a path to following directory structure:\n path\n folders for picture classification I.E. the way you want to classify pictures: folder names are classes\n folders for object classification I.E. Object_finders: folder names are main_categories\n folders for further object classification inside an Object_finder: folder names are sub_categories\n picture files for model training\n")

path = input()

print("To make picture classification work better, you may want object detectors to be avare of other object detectors. \nThis way multiple object detectors doesn't classify same objects, however training takes a lot more time. \nDo you want detectors be aware of others? If yes, give 'Y', else 'N'.\n")

impertinent = input()

if impertinent == "Y":

impertinent = True

else:

impertinent = False

print("Do you want to save old trained object detectors and not recreate them? \nIf yes, give 'Y', else 'N' and if you want to quit, give 'Q'.\n")

char = input()

if char == "Y":

if pic_cla == None:

pic_cla = Picture_classifier.deserialize()

if pic_cla == None:

print("Cannot deserialize model. Creating completely new model.\n")

pic_cla = Picture_classifier(path, impertinent)

else:

pic_cla.train_picture_classifier_model(path, False, impertinent)

elif char == "N":

if pic_cla == None:

pic_cla = Picture_classifier(path, impertinent)

else:

pic_cla.train_picture_classifier_model(path, True, impertinent)

else:

break

elif (char == "S"):

print("Write the path of the root directory, where are pictures to be sorted.\n")

input_path = input()

print("Write the path of the root directory, where sorted pictures will be moved.\n")

output_path = input()

if pic_cla == None:

pic_cla = Picture_classifier.deserialize()

if pic_cla == None:

print("Cannot deserialize model. You need to train new model.\n")

break

pic_cla.classify_pictures(input_path, output_path)

elif (char == "Q"):

break

if __name__ == "__main__":

main()