Edmond Yoo
2018-08-30 e3f280fde2fe295477aff41c31ccedd0a4ef4e3a
Visibility criteria for objects
1 files modified
40 ■■■■ changed files
transform_data.py 40 ●●●● patch | view | raw | blame | history
transform_data.py
@@ -6,6 +6,7 @@
import pandas as pd
import fetch_data
import generate_data
from shapely import geometry
card_mask = cv2.imread('data/mask.png')
@@ -50,6 +51,7 @@
        Display the current state of the generator
        :return: none
        """
        self.check_visibility()
        img_bg = cv2.resize(self.img_bg, (self.width, self.height))
        for card in self.cards:
@@ -84,11 +86,12 @@
            img_bg_crop = np.where(img_card_crop, img_card_crop, img_bg_crop)
            img_bg[bg_crop_y1:bg_crop_y2, bg_crop_x1:bg_crop_x2] = img_bg_crop
            #for extracted_object in card.objects:
            #    for pt in extracted_object.key_pts:
            #        cv2.circle(img_bg, card.coordinate_in_generator(pt[0], pt[1]), 2, (0, 0, 255), 2)
            #    bounding_box = card.bb_in_generator(extracted_object.key_pts)
            #    cv2.rectangle(img_bg, bounding_box[0], bounding_box[2], (0, 255, 0), 2)
            for ext_obj in card.objects:
                if ext_obj.visible:
                    for pt in ext_obj.key_pts:
                        cv2.circle(img_bg, card.coordinate_in_generator(pt[0], pt[1]), 2, (0, 0, 255), 2)
                    bounding_box = card.bb_in_generator(ext_obj.key_pts)
                    cv2.rectangle(img_bg, bounding_box[0], bounding_box[2], (0, 255, 0), 2)
        cv2.imshow('Result', img_bg)
        cv2.waitKey(0)
@@ -115,6 +118,30 @@
        """
        pass
    def check_visibility(self, visibility=0.5):
        """
        Check whether if extracted objects in each card are visible in the current scenario, and update their status
        :param visibility: minimum ratio of the object's area that aren't covered by another card to be visible
        :return: none
        """
        card_poly_list = [geometry.Polygon([card.coordinate_in_generator(0, 0),
                                            card.coordinate_in_generator(0, len(card.img)),
                                            card.coordinate_in_generator(len(card.img[0]), len(card.img)),
                                            card.coordinate_in_generator(len(card.img[0]), 0)]) for card in self.cards]
        # First card in the list is overlaid on the bottom of the card pile
        for i in range(len(self.cards)):
            card = self.cards[i]
            for ext_obj in card.objects:
                obj_poly = geometry.Polygon([card.coordinate_in_generator(pt[0], pt[1]) for pt in ext_obj.key_pts])
                obj_area = obj_poly.area
                # Check if the other cards are blocking this object
                for card_poly in card_poly_list[i + 1:]:
                    obj_poly = obj_poly.difference(card_poly)
                visible_area = obj_poly.area
                print("%s: %.1f visible" % (ext_obj.label, visible_area / obj_area))
                ext_obj.visible = obj_area * visibility <= visible_area
    def export_training_data(self, out_dir):
        """
        Export the generated training image along with the txt file for all bounding boxes
@@ -241,6 +268,7 @@
    def __init__(self, label, key_pts):
        self.label = label
        self.key_pts = key_pts
        self.visible = False
def main():
@@ -256,7 +284,7 @@
        if not is_planeswalker:
            card_pool = card_pool.append(card_info)
    a = 1
    for i in [random.randrange(0, card_pool.shape[0] - 1, 1) for _ in range(24)]:
    for i in [random.randrange(0, card_pool.shape[0] - 1, 1) for _ in range(20)]:
        card_info = card_pool.iloc[i]
        img_name = '../usb/data/png/%s/%s_%s.png' % (card_info['set'], card_info['collector_number'],
                                                     fetch_data.get_valid_filename(card_info['name']))