Merge pull request #99 from PracticalDL/siddha/chapter-8

Add OCR

Author: meherkasam

code/chapter-8/experiment-scripts/optical-character-recognition/amazon.py

@@ -0,0 +1,61 @@
+import http.client, urllib, base64
+import sys
+import os
+from os import listdir
+from os.path import isfile, join
+import json
+try:
+ from urllib.parse import urlparse
+except ImportError:
+ from urlparse import urlparse
+import boto3
+
+def amazon(filename):
+ with open(filename, "rb") as image_file:
+ image_bytes = image_file.read()
+ client = boto3.client('rekognition')
+ response = client.detect_text(Image={'Bytes': image_bytes })
+ text_detections = response[u'TextDetections']
+ result = []
+ for each in text_detections:
+ result.append(each[u'DetectedText'])
+ return result
+
+def getopts(argv):
+ opts = {} # Empty dictionary to store key-value pairs.
+ while argv: # While there are arguments left to parse.
+ if argv[0][0] == '-': # Found a "-name value" pair.
+ opts[argv[0]] = argv[1] # Add key and value to the dictionary.
+ argv = argv[1:] # Reduce the argument list by copying it starting from index 1.
+ return opts
+
+def process_image(image):
+ results = {}
+ results[image] = amazon(image)
+ return results
+
+def process_images(directory):
+ images = [join(directory, f) for f in listdir(directory) if isfile(join(directory, f)) and (f.endswith(".jpg") or f.endswith(".jpeg") or f.endswith(".png"))]
+ results = {}
+ for image in images:
+ results[image] = amazon(image)
+ return results
+
+if __name__ == '__main__':
+ from sys import argv
+ input_args = getopts(argv)
+ results = []
+ if '-i' in input_args and '-o' in input_args:
+ image = input_args['-i']
+ results = process_image(image)
+ output_path = input_args['-o']
+ with open(output_path, "w") as write_file:
+ json.dump(results, write_file)
+ elif '-d' in input_args and '-o' in input_args:
+ directory = input_args['-d']
+ results = process_images(directory)
+ output_path = input_args['-o']
+ with open(output_path, "w") as write_file:
+ json.dump(results, write_file)
+ else:
+ print("Usage: python amazon.py [-i path to an image | -d path to directory of images] [-o output path]")

code/chapter-8/experiment-scripts/optical-character-recognition/google.py

@@ -0,0 +1,90 @@
+import http.client, urllib, base64
+import sys
+import os
+from os import listdir
+from os.path import isfile, join
+import json
+try:
+ from urllib.parse import urlparse
+except ImportError:
+ from urlparse import urlparse
+import boto3
+
+google_api_key = 'ADD_YOUR_KEY_HERE'
+
+def google(filename):
+ with open(filename, "rb") as image_file:
+ encoded_string = base64.b64encode(image_file.read())
+
+ endpoint = "/v1/images:annotate?key=" + google_api_key
+ 
+ detection_type = "TEXT_DETECTION"
+ request_body = {
+ "requests":[
+ {
+ "image":{
+ "content":encoded_string
+ },
+ "features":[
+ {
+ "type":detection_type,
+ "maxResults":10
+ }
+ ]
+ }
+ ]
+ }
+
+ try:
+ conn = http.client.HTTPSConnection('vision.googleapis.com')
+ conn.request("POST", endpoint, json.dumps(request_body))
+ response = conn.getresponse()
+ data = response.read()
+ json_data = json.loads(data)
+ result = []
+ annotations = json_data['responses'][0]['textAnnotations']
+ for annotation in annotations:
+ result.append(annotation['description'])
+ return result
+ conn.close()
+ except Exception as e:
+ print(e)
+
+def getopts(argv):
+ opts = {} # Empty dictionary to store key-value pairs.
+ while argv: # While there are arguments left to parse.
+ if argv[0][0] == '-': # Found a "-name value" pair.
+ opts[argv[0]] = argv[1] # Add key and value to the dictionary.
+ argv = argv[1:] # Reduce the argument list by copying it starting from index 1.
+ return opts
+
+def process_image(image):
+ results = {}
+ results[image] = google(image)
+ return results
+
+def process_images(directory):
+ images = [join(directory, f) for f in listdir(directory) if isfile(join(directory, f)) and (f.endswith(".jpg") or f.endswith(".jpeg") or f.endswith(".png"))]
+ results = {}
+ for image in images:
+results[image] = google(image)
+ return results
+
+
+if __name__ == '__main__':
+ input_args = getopts(sys.argv)
+ results = []
+ if '-i' in input_args and '-o' in input_args:
+ image = input_args['-i']
+ results = process_image(image)
+ output_path = input_args['-o']
+ with open(output_path, "w") as write_file:
+ json.dump(results, write_file)
+ elif '-d' in input_args and '-o' in input_args:
+ directory = input_args['-d']
+ results = process_images(directory)
+ output_path = input_args['-o']
+ with open(output_path, "w") as write_file:
+ json.dump(results, write_file)
+ else:
+ print("Usage: python google.py [-i path to an image | -d path to directory of images] [-o output path]")

code/chapter-8/experiment-scripts/optical-character-recognition/microsoft-phase-1.py

@@ -0,0 +1,74 @@
+import http.client, urllib
+from sys import argv
+import json
+import os
+from os import listdir
+from os.path import isfile, join
+try:
+ from urllib.parse import urlparse
+except ImportError:
+ from urlparse import urlparse
+
+microsoft_api_key = 'ADD_YOUR_KEY_HERE'
+path_to_images = "http://ADD_YOUR_IP_ADDRESS_HERE/images/"
+
+def microsoft_phase_1(filename):
+ headers = {
+ 'Content-Type': 'application/json',
+ 'Ocp-Apim-Subscription-Key': microsoft_api_key,
+ }
+
+ endpoint = "/vision/v2.0/recognizeText?mode=Printed"
+ params = urllib.urlencode({
+ # Request parameters
+ # 'mode': 'Printed'
+ })
+ try:
+ conn = http.client.HTTPSConnection('westus.api.cognitive.microsoft.com')
+ request_body = "{\"url\":\"" + path_to_images + filename + "\"}"
+ conn.request("POST", endpoint, request_body, headers)
+ response = conn.getresponse()
+ url = response.getheader("Operation-Location")
+ recognition_id = urlparse(url).path.split('/')[-1]
+ return recognition_id
+ conn.close()
+ except Exception as e:
+ print(e) 
+ print("Error")
+
+
+def getopts(argv):
+ opts = {} # Empty dictionary to store key-value pairs.
+ while argv: # While there are arguments left to parse.
+ if argv[0][0] == '-': # Found a "-name value" pair.
+ opts[argv[0]] = argv[1] # Add key and value to the dictionary.
+ argv = argv[1:] # Reduce the argument list by copying it starting from index 1.
+ return opts
+
+def process_image(image):
+ results = {}
+ results[image] = microsoft_phase_1(image)
+ return results
+
+def process_images(directory):
+ results = {}
+ images = [f for f in listdir(directory) if isfile(join(directory, f))]
+ for image in images:
+ results[image] = microsoft_phase_1(image)
+ return results
+
+if __name__ == '__main__':
+ input_args = getopts(argv)
+ results = []
+ if '-i' in input_args:
+ image = input_args['-i']
+ results = process_image(image)
+ for key, value in results.iteritems():
+ print(key + "," + value)
+ elif '-d' in input_args:
+ directory = input_args['-d']
+ results = process_images(directory)
+ for key, value in results.iteritems():
+ print(key + "," + value)
+ else:
+ print("Usage: python microsoft-phase-1.py [-i path to an image | -d path to directory of images]")

code/chapter-8/experiment-scripts/optical-character-recognition/microsoft-phase-2.py

@@ -0,0 +1,61 @@
+import http.client, urllib, base64
+import sys
+import os
+from os import listdir
+from os.path import isfile, join
+import json
+import base64
+try:
+ from urllib.parse import urlparse
+except ImportError:
+ from urlparse import urlparse
+import boto3
+
+microsoft_api_key = 'ADD_YOUR_KEY_HERE'
+
+def microsoft_phase_2(recognition_id):
+ headers = {
+ 'Ocp-Apim-Subscription-Key': microsoft_api_key,
+ }
+
+ try:
+ conn = http.client.HTTPSConnection('westus.api.cognitive.microsoft.com')
+ conn.request("GET", "/vision/v2.0/textOperations/" + recognition_id, "", headers)
+ result = []
+ response = conn.getresponse()
+ return response.read()
+ except Exception as e:
+ print(e)
+
+def getopts(argv):
+ opts = {} # Empty dictionary to store key-value pairs.
+ while argv: # While there are arguments left to parse.
+ if argv[0][0] == '-': # Found a "-name value" pair.
+ opts[argv[0]] = argv[1] # Add key and value to the dictionary.
+ argv = argv[1:] # Reduce the argument list by copying it starting from index 1.
+ return opts
+
+def process_images(recognition_id_filename):
+ data = []
+ with open(recognition_id_filename) as file:
+ data = file.readlines()
+ results = {}
+ for each in data:
+ image_name = each.split(",")[0].strip()
+ recognition_id = each.split(",")[-1].strip("\n").strip()
+ results[image_name] = microsoft_phase_2(recognition_id)
+ return results
+
+
+if __name__ == '__main__':
+ from sys import argv
+ input_args = getopts(argv)
+ results = []
+ if '-i' in input_args and '-o' in input_args:
+ recognition_ids_filename = input_args['-i']
+ results = process_images(recognition_ids_filename)
+ output_path = input_args['-o']
+ with open(output_path, "w") as write_file:
+ json.dump(results, write_file)
+ else:
+ print("Usage: python microsoft-phase-2.py [-i path to recognition IDs file] [-o output path]")