Graphical Photo Labeling Create a new Python project in PyCharm and add the following files to it: image_helpers.pyPreview the documentView in a new window labels_graphical.pyPreview the documentView in a new window In labels_graphical.py, there are a number of (helper) functions written for you. Then near the bottom of the file, you will fill in the code in the get_labels and label_image functions, per the requirements below (and in the functions' docstrings). get_labels has 1 required parameter, img, which is either the image bytes or a string that is the URL or filename for an image. The second parameter is an optional one, that is the confidence level used when getting the labels from AWS Rekognition. get_labels will contact AWS Rekognition to get the labels for the image and return the labels and their confidence levels. image_helpers.py below: import requests def get_image_from_url(imgurl): """ Loads and returns the bytes of the image from the specified url :param imgurl: the url """ resp = requests.get(imgurl) imgbytes = resp.content return imgbytes def get_image_from_file(filename): """ Loads and returns the bytes of the image from the specified file :param filename: the name of the file Based on https://docs.aws.amazon.com/rekognition/latest/dg/example4.html, last access 10/3/2017 """ with open(filename, 'rb') as imgfile: return imgfile.read() def get_image(img): """ Loads and returns the image either from a URL or a file :param img: string that is either the URL or file :return: """ if img.lower().startswith('http'): return get_image_from_url(img) else: return get_image_from_file(img) labels_graphical.py below: from PIL import Image, ImageDraw, ImageFont import boto3 from pprint import pprint, pformat from io import BytesIO import image_helpers # -------------------------------------------------------------------- # DO NOT CHANGE THESE FUNCTIONS def format_text(text, columns): ''' Returns a copy of text that will not span more than the specified number of columns :param text: the text :param columns: the maximum number of columns :return: the formatted text ''' # format the text to fit the specified columns import re text = re.sub('[()\']', '', pformat(text, width=columns)) text = re.sub(' ', ' ', text) return text def text_rect_size(draw, text, font): ''' Returns the size of the rectangle to be used to draw as the background for the text :param draw: an ImageDraw.Draw object :param text: the text to be displayed :param font: the font to be used :return: the size of the rectangle to be used to draw as the background for the text ''' (width, height) = draw.multiline_textsize(text, font=font) return (width * 1.1, height * 1.3) def add_text_to_img(img, text, pos=(0, 0), color=(0, 0, 0), bgcolor=(255, 255, 255, 128), columns=60, font=ImageFont.truetype('ariblk.ttf', 22)): ''' Creates and returns a copy of the image with the specified text displayed on it :param img: the (Pillow) image :param text: the text to display :param pos: a 2 tuple containing the xpos, and ypos of the text :param color: the fill color of the text :param bgcolor: the background color of the box behind the text :param columns: the max number of columns for the text :param font: the font to use :return: a copy of the image with the specified text displayed on it ''' # make a blank image for the text, initialized to transparent text color txt_img = Image.new('RGBA', img.size, (255, 255, 255, 0)) draw = ImageDraw.Draw(txt_img) # format the text text = format_text(text, columns) # get the size of the text drawn in the specified font (text_width, text_height) = ImageDraw.Draw(img).multiline_textsize(text, font=font) # compute positions and box size (xpos, ypos) = pos rwidth = text_width * 1.1 rheight = text_height * 1.4 text_xpos = xpos + (rwidth - text_width) / 2 text_ypos = ypos + (rheight - text_height) / 2 # draw the rectangle (slightly larger) than the text draw.rectangle([xpos, ypos, xpos + rwidth, ypos + rheight], fill=bgcolor) # draw the text on top of the rectangle draw.multiline_text((text_xpos, text_ypos), text, font=font, fill=color) del draw # clean up the ImageDraw object return Image.alpha_composite(img.convert('RGBA'), txt_img) def get_pillow_img(imgbytes): """ Creates and returns a Pillow image from the given image bytes :param imgbytes: the bytes of the image """ return Image.open(BytesIO(imgbytes)) # END DO NOT CHANGE SECTION # -------------------------------------------------------------------- def get_labels(img, confidence=50): """ Gets the labels from AWS Rekognition for the given image :param img: either the image bytes or a string that is the URL or filename for an image :param confidence: the confidence level (defaults to 50) >>> get_labels('http://www.idothat.us/images/idothat- img/features/pool-patio-lanai/ft-pool-patio-lanai-2.jpg') [{'Name': 'Pool', 'Confidence': 96.66264343261719}, {'Name': 'Water', 'Confidence': 96.66264343261719}, {'Name': 'Resort', 'Confidence': 76.01207733154297}, {'Name': 'Swimming Pool', 'Confidence': 76.01207733154297}, {'Name': 'Balcony', 'Confidence': 66.56996154785156}, {'Name': 'Patio', 'Confidence': 62.89345169067383}, {'Name': 'Flagstone', 'Confidence': 59.46694564819336}, {'Name': 'Backyard', 'Confidence': 56.5162467956543}, {'Name': 'Yard', 'Confidence': 56.5162467956543}, {'Name': 'Path', 'Confidence': 52.408050537109375}, {'Name': 'Sidewalk', 'Confidence': 52.408050537109375}, {'Name': 'Walkway', 'Confidence': 52.408050537109375}, {'Name': 'Alley', 'Confidence': 50.79618835449219}, {'Name': 'Alleyway', 'Confidence': 50.79618835449219}, {'Name': 'Road', 'Confidence': 50.79618835449219}, {'Name': 'Street', 'Confidence': 50.79618835449219}, {'Name': 'Town', 'Confidence': 50.79618835449219}, {'Name': 'Plant', 'Confidence': 50.6763801574707}] >>> get_labels('http://www.idothat.us/images/idothat- img/features/pool-patio-lanai/ft-pool-patio-lanai-2.jpg', 90) [{'Name': 'Pool', 'Confidence': 96.66264343261719}, {'Name': 'Water', 'Confidence': 96.66264343261719}] """ # replace pass below with your implementation pass def label_image(img, confidence=50): ''' Creates and returns a copy of the image, with labels from Rekognition displayed on it :param img: a string that is either the URL or filename for the image :param confidence: the confidence level (defaults to 50) :return: a copy of the image, with labels from Rekognition displayed on it ''' # replace pass below with your implementation pass if __name__ == "__main__": # can't use input since PyCharm's console causes problems entering URLs # img = input('Enter either a URL or filename for an image: ') img = 'https://blog.njsnet.co/content/images/2017/02/trumprecognition.png' img = 'https://www.parrots.org/images/uploads/dreamstime_C_47716185.jpg' labelled_image = label_image(img) labelled_image.show()