data_tool.py

import config
from tool import img_path_to_GEI
import logging
import os
logger = logging.getLogger("data")


def load_training_validation_data(train_view=None, val_view=None, train_dir=None, val_dir=None):
    human_id = ["%03d" % i for i in range(1, 125)]
    if train_dir is None:
        train_dir = ["nm-%02d" % i for i in range(1, 5)]
    if val_dir is None:
        val_dir = ["nm-05", "nm-06"]
    if train_view is None:
        train_view = "090"
    if val_view is None:
        val_view = "090"

    training_x = []
    training_y = []

    validation_x = []
    validation_y = []

    # check dir exists
    for id in human_id:
        for dir in train_dir:
            img_dir = "%s/%s/%s/%s" % (config.Project.casia_dataset_b_path, id, dir, train_view)
            if not os.path.exists(img_dir):
                logger.warning("%s do not exist" % img_dir)
        for dir in val_dir:
            img_dir = "%s/%s/%s/%s" % (config.Project.casia_dataset_b_path, id, dir, val_view)
            if not os.path.exists(img_dir):
                logger.warning("%s do not exist" % img_dir)

    for id in human_id:
        logger.info("processing human %s" % id)
        for dir in train_dir:
            img_dir = "%s/%s/%s/%s" % (config.Project.casia_dataset_b_path, id, dir, train_view)
            data = img_path_to_GEI(img_dir)

            if len(data.shape) > 0:
                training_x.append(data)
                training_y.append(id)
            else:
                logger.warning("fail to extract %s of %s" % (img_dir, id))

        for dir in val_dir:
            img_dir = "%s/%s/%s/%s" % (config.Project.casia_dataset_b_path, id, dir, val_view)
            data = img_path_to_GEI(img_dir)
            if len(data.shape) > 0:
                validation_x.append(data)
                validation_y.append(id)
            else:
                logger.warning("fail to extract %s of %s" % (img_dir, id))

    return training_x, training_y, validation_x, validation_y


def output_result(view_list, correct_tbl):
    logger.info("every row means the validation result from different training views")
    logger.info("\t\t" + "\t\t".join(view_list))
    for val_view in view_list:
        output = "%s\t" % val_view
        for train_view in view_list:
            key = "%s-%s" % (train_view, val_view)
            if key in correct_tbl:
                precision = correct_tbl[key]
            else:
                precision = 0.0
            output += "%.2f\t" % precision
        logger.info(output)

if __name__ == '__main__':
    level = logging.INFO
    FORMAT = '%(asctime)-12s[%(levelname)s] %(message)s'
    logging.basicConfig(level=level, format=FORMAT, datefmt='%Y-%m-%d %H:%M:%S')

    view_list = ["%03d" % x for x in range(0, 181, 18)]
    correct_tbl = {}
    for train_view in view_list:
        for val_view in view_list:
            correct_tbl["%s-%s" % (train_view, val_view)] = 0
    output_result(view_list, correct_tbl)

    # training_x, training_y, validation_x, validation_y = load_training_validation_data()
    # count = 0
    # for x in training_x:
    #     count += 1
    #     imsave("%s/%03d.bmp" % (config.project.test_data_path, count), x)