import argparse def get_args(): parser = argparse.ArgumentParser(description='Process some integers.') parser.add_argument('--model_dir', type = str, help =' the directory of the model,default using the Bird_D model included', default='') parser.add_argument('--det_model', type = str, help ='you can select from yolo5,fasterrcnn, retinanetknn, retinanet, megadetector and yolonas', default='retinanet') parser.add_argument('--det_conf', type = float, default=0.1, help ='Confidence threshold of your detection model') parser.add_argument('--cla_model', type = str, help ='you can select from res18 and mixmatch', default='') parser.add_argument('--image_root',type = str, help = 'The root dir where image are stores') parser.add_argument('--csv_root',type = str, default='', help = 'The root dir where image info are stores') parser.add_argument('--image_ext',type = str, default = 'JPG', help = 'the extension of the image(without dot), default is JPG') parser.add_argument('--image_altitude',type = int, default = 90, help = 'the altitude of the taken image, default is set to be 90') parser.add_argument('--image_location',type = str, default = 'No_Where', help = 'the location of the taken image, default is set to be 90') parser.add_argument('--image_date',type = str, default = '2022-10-26', help = 'the date of the taken image, default is set to be 2022-10-26') parser.add_argument('--use_altitude',type = bool, default = True, help = 'whether to use altitude to scale the image, default is True') parser.add_argument('--out_dir',type = str, help = 'where the output will be generated,default is ./results', default = './results') parser.add_argument('--visualize',type = bool, help = 'whether to have visualization stored to result, default is True', default = False) parser.add_argument('--evaluate',type = bool, help = 'whether to evaluate the reslt,default is False', default = False) args = parser.parse_args() #if the image_root input is with extension(*.JPG) wrap into list #else fetch the list of image return args