def shot_detection(path,csvpath):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.SHOT_CHANGE_DETECTION]
with io.open(path, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(
features=features, input_content=input_content)
print('\nProcessing video for shot change annotations:',path)
result = operation.result(timeout=200)
print('\nFinished processing.')
fields = ['Shot Number', 'Start', 'End']
rows=[]
# first result is retrieved because a single video was processed
for i, shot in enumerate(result.annotation_results[0].shot_annotations):
start_time = (shot.start_time_offset.seconds +
shot.start_time_offset.nanos / 1e9)
end_time = (shot.end_time_offset.seconds +
shot.end_time_offset.nanos / 1e9)
rows.append([i+1,start_time,end_time])
# writing to csv file
with open(csvpath, 'w') as csvfile:
# creating a csv writer object
csvwriter = csv.writer(csvfile)
# writing the fields
csvwriter.writerow(fields)
# writing the data rows
csvwriter.writerows(rows)
def analyze_shots(path):
""" Detects camera shot changes. """
# [START video_shot_tutorial_construct_request]
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.Feature.SHOT_CHANGE_DETECTION]
operation = video_client.annotate_video(
request={"features": features, "input_uri": path}
)
# [END video_shot_tutorial_construct_request]
print("\nProcessing video for shot change annotations:")
# [START video_shot_tutorial_check_operation]
result = operation.result(timeout=120)
print("\nFinished processing.")
# [END video_shot_tutorial_check_operation]
# [START video_shot_tutorial_parse_response]
for i, shot in enumerate(result.annotation_results[0].shot_annotations):
start_time = (
shot.start_time_offset.seconds + shot.start_time_offset.microseconds / 1e6
)
end_time = (
shot.end_time_offset.seconds + shot.end_time_offset.microseconds / 1e6
)
print("\tShot {}: {} to {}".format(i, start_time, end_time))
def performAllAnalysis(self, path):
myLogger.debug(
"Entering processing gs stored video for label annotations" + path)
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
myLogger.debug("LABEL_DETECTION initiated")
features += [
videointelligence.enums.Feature.EXPLICIT_CONTENT_DETECTION
]
myLogger.debug("EXPLICIT_CONTENT_DETECTION initiated")
features += [videointelligence.enums.Feature.SHOT_CHANGE_DETECTION]
myLogger.debug("SHOT_CHANGE_DETECTION initiated")
# features += [videointelligence.enums.Feature.FACE_DETECTION]
# print(features)
# exit
mode = videointelligence.enums.LabelDetectionMode.SHOT_AND_FRAME_MODE
config = videointelligence.types.LabelDetectionConfig(
label_detection_mode=mode)
myLogger.debug("SHOT_AND_FRAME_MODE set")
context = videointelligence.types.VideoContext(
label_detection_config=config)
operation = video_client.annotate_video(path,
features=features,
video_context=context)
myLogger.debug("Start video analysis for: " + path)
result = operation.result(timeout=3600)
myLogger.debug("Finished video analysis for: " + path)
return result
def transcribe(path):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.SPEECH_TRANSCRIPTION]
config = videointelligence.types.SpeechTranscriptionConfig(
language_code="en-US", enable_automatic_punctuation=True)
video_context = videointelligence.types.VideoContext(
speech_transcription_config=config)
operation = video_client.annotate_video(input_uri=path,
features=features,
video_context=video_context)
print("\nProcessing video for speech transcription.")
result = operation.result(timeout=600)
content = ""
# There is only one annotation_result since only
# one video is processed.
annotation_results = result.annotation_results[0]
# The number of alternatives for each transcription is limited by
# SpeechTranscriptionConfig.max_alternatives.
# Each alternative is a different possible transcription
# and has its own confidence score.
# for alternative in speech_transcription.alternatives:
# content += alternative.transcript
# Returns whole transcript
return "".join(annotation_results.speech_transcriptions)
def analyze_video_labels(path):
""" Detects labels given a GCS path. """
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with open(path, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(features=features,
input_content=input_content)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
confidence = segment.confidence
# print('\tSegment {}: {}'.format(i, positions))
print('\tConfidence: {}'.format(confidence))
print('\n')
def analyze_labels(path):
""" Detects labels given a GCS path. """
# [START video_label_tutorial_construct_request]
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(input_uri=path, features=features)
# [END video_label_tutorial_construct_request]
print('\nProcessing video for label annotations:')
# [START video_label_tutorial_check_operation]
result = operation.result(timeout=90)
print('\nFinished processing.')
# [END video_label_tutorial_check_operation]
# [START video_label_tutorial_parse_response]
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
confidence = segment.confidence
print('\tSegment {}: {}'.format(i, positions))
print('\tConfidence: {}'.format(confidence))
print('\n')
def run_quickstart():
# [START video_quickstart]
from google.cloud import videointelligence
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video('gs://demomaker/cat.mp4',
features=features)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
print('\nFinished processing.')
# first result is retrieved because a single video was processed
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
confidence = segment.confidence
print('\tSegment {}: {}'.format(i, positions))
print('\tConfidence: {}'.format(confidence))
print('\n')
def __init__(self, logger, event):
#variables
self.logger = logger.global_log
folder_temporary = 'C:/Temp'
video_content = event['video-content']
credentials = "syn-g-cloud-ac072cf6a455.json"
input_content = ''
location = 'us-east1'
#build
tmp_video = '{}/{}'.format(folder_temporary, video_content)
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = credentials
gcp_feats = videointelligence.enums.Feature
#open
with io.open(tmp_video, 'rb') as file:
input_content = file.read()
#init the service
video_client = videointelligence.VideoIntelligenceServiceClient()
#feats to be called
features = [gcp_feats.LABEL_DETECTION, gcp_feats.SHOT_CHANGE_DETECTION]
#call
operation = video_client.annotate_video(input_content=input_content,
features=features,
location_id=location)
#result
result = operation.result(timeout=900)
#print
self.get_scenes(result)
self.get_labels_segment(result)
self.get_labels_shot(result)
def run_quickstart():
# [START video_quickstart]
from google.cloud import videointelligence
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(
request={
"features": features,
"input_uri": "gs://cloud-samples-data/video/cat.mp4",
})
print("\nProcessing video for label annotations:")
result = operation.result(timeout=120)
print("\nFinished processing.")
# first result is retrieved because a single video was processed
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print("Video label description: {}".format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print("\tLabel category description: {}".format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.microseconds / 1e6)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.microseconds / 1e6)
positions = "{}s to {}s".format(start_time, end_time)
confidence = segment.confidence
print("\tSegment {}: {}".format(i, positions))
print("\tConfidence: {}".format(confidence))
print("\n")
def google(screen_name):
#https://cloud.google.com/video-intelligence/docs/analyze-labels#video_analyze_labels-python
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with io.open(path, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(features=features,
input_content=input_content)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
print('\nFinished processing.')
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
confidence = segment.confidence
print('\tConfidence: {}'.format(confidence))
print('\n')
def analyzeLabels(input_content):
videoClient = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = videoClient.annotate_video(input_content=input_content,
features=features)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
print('\nFinished processing.')
shotLabels = result.annotation_results[0].shot_label_annotations
for i, shot_label in enumerate(shotLabels):
print('Shot label description: {}'.format(
shot_label.entity.description))
for category_entity in shot_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, shot in enumerate(shot_label.segments):
startTime = (shot.segment.start_time_offset.seconds +
shot.segment.start_time_offset.nanos / 1e9)
endTime = (shot.segment.end_time_offset.seconds +
shot.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(startTime, endTime)
confidence = shot.confidence
print('\tSegment {}: {}'.format(i, positions))
print('\tConfidence: {}'.format(confidence))
print('\n')
def analyze_faces(path):
# [START construct_request]
""" Detects faces given a GCS path. """
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.FACE_DETECTION]
operation = video_client.annotate_video(path, features=features)
# [END construct_request]
print('\nProcessing video for face annotations:')
# [START check_operation]
result = operation.result(timeout=600)
print('\nFinished processing.')
# [END check_operation]
# [START parse_response]
# first result is retrieved because a single video was processed
faces = result.annotation_results[0].face_annotations
for face_id, face in enumerate(faces):
print('Thumbnail size: {}'.format(len(face.thumbnail)))
for segment_id, segment in enumerate(face.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
print('\tSegment {}: {}'.format(segment_id, positions))
def performLabelAnalysis(self, path):
""" Detects labels given a GCS path. """
myLogger.debug(
"Entering label processing gs stored video for label annotations" +
path)
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
# features += [videointelligence.enums.Feature.FACE_DETECTION]
# print(features)
# exit
mode = videointelligence.enums.LabelDetectionMode.SHOT_AND_FRAME_MODE
config = videointelligence.types.LabelDetectionConfig(
label_detection_mode=mode)
context = videointelligence.types.VideoContext(
label_detection_config=config)
myLogger.debug("SHOT_AND_FRAME_MODE set")
operation = video_client.annotate_video(path,
features=features,
video_context=context)
myLogger.debug("SHOT_AND_FRAME_MODE set")
myLogger.debug("Start label video analysis for: " + path)
result = operation.result(timeout=3600)
myLogger.debug("Finished label video analysis for: " + path)
return result
def create_video_speech_operation_from_model(
self, storage_uri, *args,
**kwargs): # i.e. copied and pasted from Google
"""Transcribe speech from a video stored on GCS."""
from google.cloud import videointelligence
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.Feature.SPEECH_TRANSCRIPTION]
config = videointelligence.SpeechTranscriptionConfig(
language_code="en-US",
enable_automatic_punctuation=True,
#enable_word_time_offsets=True,
max_alternatives=2)
video_context = videointelligence.VideoContext(
speech_transcription_config=config)
operation = video_client.annotate_video(
request={
"features": features,
"input_uri": storage_uri,
"video_context": video_context,
})
# Use confirmation
if self.require_api_confirmation:
confirmation = input(
f"Really recognize speech in {storage_uri}? (Y/n) ")
if confirmation.lower() != "y":
raise Exception("Did not agree to recognize speech")
print("\nProcessing video for speech transcription.")
return operation
def zoom_analysis(video, frame_to_use=0):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(video, features=features)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=120)
print('\nFinished processing.')
frame_offsets = []
# first result is retrieved because a single video was processed
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
for category_entity in segment_label.category_entities:
# Take frames with people
if (category_entity.description == 'person'):
print('\tLabel category description: {}'.format(
category_entity.description))
print(segment_label)
frame = segment_label.frames[frame_to_use]
time_offset = (frame.time_offset.seconds +
frame.time_offset.nanos / 1e9)
print('\tFirst frame time offset: {}s'.format(time_offset))
print('\tFirst frame confidence: {}'.format(frame.confidence))
print('\n')
frame_offsets.append(time_offset)
return sorted(set(frame_offsets))
def process_audio(file, context):
bucket = validate_message(file, "bucket")
name = validate_message(file, "name")
print("Starting transcription for", bucket, " / ", name)
video_client = videointelligence.VideoIntelligenceServiceClient()
# maybe add other features here?
features = [videointelligence.enums.Feature.SPEECH_TRANSCRIPTION]
config = videointelligence.types.SpeechTranscriptionConfig(
language_code="en-US", enable_automatic_punctuation=True)
video_context = videointelligence.types.VideoContext(
speech_transcription_config=config)
operation = video_client.annotate_video(input_uri=f"gs://{bucket}/{name}",
features=features,
video_context=video_context)
transcriptions = []
annotation_results = operation.result()
first_result = annotation_results.annotation_results[0]
speech_transcriptions_raw = first_result.speech_transcriptions
for result in speech_transcriptions_raw:
alternative = result.alternatives[0]
transcript = alternative.transcript
confidence = alternative.confidence
words = []
for word_info in alternative.words:
word = word_info.word
start = word_info.start_time.seconds + 1e-9 * word_info.start_time.nanos
end = word_info.end_time.seconds + 1e-9 * word_info.end_time.nanos
words.append({"word": word, "start": start, "end": end})
keywords = text_to_keywords.get_keywords(transcript)
transcriptions.append({
"transcript": transcript,
"confidence": confidence,
"words": words,
"keywords": keywords
})
# response = transcribe_file(output, channels)
# response = [paragraph.to_json() for paragraph in response]
mongo.set_keyword_data(file["name"], transcriptions)
print("File {} processed.".format(file["name"]))
return transcriptions
# process_audio({"bucket": "catchup-app", "name": "test_video.mp4"}, {})
def google_analyze(self):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with io.open(self.name + '.mp4', 'rb') as movie:
input_content = movie.read()
try:
operation = video_client.annotate_video(
features=features, input_content=input_content)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
except Exception as e:
print("Video Intelligence error")
exit()
print('\nFinished processing.')
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print("Label category description: " +
category_entity.description)
for i, segment in enumerate(segment_label.segments):
confidence = segment.confidence
print("The accuracy of the identification in this case is " +
str(confidence) + "\n")
def google_analysis():
#https://cloud.google.com/video-intelligence/docs/analyze-labels#video_analyze_labels-python
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = os.getcwd() + '/google.json'
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with io.open(os.path.join(os.getcwd(), "" + screen_name + ".mp4"),
'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(features=features,
input_content=input_content)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
print('\nFinished processing.')
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
#show the matching degree
for i, segment in enumerate(segment_label.segments):
confidence = segment.confidence
print('\tConfidence: {}'.format(confidence))
print('\n')
def hello():
video_file_path = request.args.get('video_file_path')
if video_file_path == None or len(video_file_path) == 0:
logging.info('no path parameter provided')
# print('no path parameter provided')
return 'no path parameter provided'
path = 'gs://' + STORAGE_NAME + '/' + video_file_path
logging.info('Accessing video from ' + path + ' ...')
# print('Accessing video from ' + path + ' ...')
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
mode = videointelligence.enums.LabelDetectionMode.SHOT_AND_FRAME_MODE
config = videointelligence.types.LabelDetectionConfig(label_detection_mode=mode)
context = videointelligence.types.VideoContext(label_detection_config=config)
operation = video_client.annotate_video(path, features=features, video_context=context)
logging.info('Processing video for labelling...')
# print('Processing video for labelling...')
result = operation.result(timeout=180)
labels = []
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
labels.append(segment_label.entity.description)
# Process shot level label annotations
shot_labels = result.annotation_results[0].shot_label_annotations
for i, shot_label in enumerate(shot_labels):
labels.append(shot_label.entity.description)
# # Process frame level label annotations
# frame_labels = result.annotation_results[0].frame_label_annotations
# for i, frame_label in enumerate(frame_labels):
# labels.append(frame_label.entity.description)
labels = list(set(labels))
logging.info(labels)
# print(labels)
if len(labels) == 0:
labels.append('please wait')
trimmed_file_path = video_file_path.split('.')[0]
ref = db.reference(trimmed_file_path)
ref.set(labels)
logging.info('Written to ' + LABELS_DATABASE_NAME)
# print('Written to ' + LABELS_DATABASE_NAME)
return jsonify(labels)
def analyze_labels_file(path, dict_list):
# [START video_analyze_labels]
"""Detect labels given a file path."""
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with io.open(path, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(
features=features, input_content=input_content)
print('\nProcessing video for label annotations (This may take a while):')
result = operation.result(timeout=300)
print('Finished processing\n')
print('****************************************')
# Process shot level label annotations
label_count = 0
print('\nGenerating a report')
shot_labels = result.annotation_results[0].shot_label_annotations
file = open('Label Analysis Result.txt', 'w')
file.write('Probable contents in the video (NOT following the picture order): \n\n')
# Make a list for MySQL database
content_name = []
for i, shot_label in enumerate(shot_labels):
content_name.append(shot_label.entity.description)
label_count += 1
file.write(' ' + str(label_count) + '\t{}'.format(
shot_label.entity.description) + '\n')
file.write('\n\n')
file.write('Details are listed below\n\n')
# Continue writing in the txt report
# And insert pic_content data to table pic_content
for i, shot_label in enumerate(shot_labels):
file.write('Shot label description: {}'.format(
shot_label.entity.description) + '\n')
for category_entity in shot_label.category_entities:
file.write('\tLabel category description: {}'.format(
category_entity.description) + '\n')
for i, shot in enumerate(shot_label.segments):
start_time = (shot.segment.start_time_offset.seconds +
shot.segment.start_time_offset.nanos / 1e9)
end_time = (shot.segment.end_time_offset.seconds +
shot.segment.end_time_offset.nanos / 1e9)
frame_no = int(round(end_time*FRAMERATE))
dict_list[frame_no-1]["contents"].append(shot_label.entity.description)
positions = '{}Picture number {}'.format('', frame_no)
confidence = shot.confidence
file.write('\tSegment {}: {}'.format(i+1, positions) + '\n')
file.write('\tConfidence: {}'.format(confidence) + '\n')
file.write('\n')
file.close()
POSTS.insert_many(dict_list)
print('Report generated\n')
print('****************************************')
def consume(self, path, **kwargs):
""" Detects labels given a GCS path. """
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
mode = videointelligence.enums.LabelDetectionMode.SHOT_AND_FRAME_MODE
config = videointelligence.types.LabelDetectionConfig(
label_detection_mode=mode)
context = videointelligence.types.VideoContext(
label_detection_config=config)
operation = video_client.annotate_video(path,
features=features,
video_context=context)
info('Processing "%s" for label annotations:' % path)
result = operation.result(timeout=10 * 60)
info('Finished processing.')
out = []
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
# print('Video label description: {}'.format(
# segment_label.entity.description))
for category_entity in segment_label.category_entities:
# print('\tLabel category description: {}'.format(
# category_entity.description))
pass
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
confidence = segment.confidence
# print('\tSegment {}: {}'.format(i, positions))
# print('\tConfidence: {}'.format(confidence))
# print('\n')
out.append({
'label':
segment_label.entity.description,
'category':
category_entity.description
if segment_label.category_entities else '',
'start':
start_time if segment_label.segments else '',
'end':
end_time if segment_label.segments else '',
'confidence':
confidence if segment_label.segments else ''
})
return out
def test_default_credentials(self):
env = EnvironmentVarGuard()
env.set('KAGGLE_USER_SECRETS_TOKEN', 'foobar')
env.set('KAGGLE_KERNEL_INTEGRATIONS', 'VIDEO_INTELLIGENCE')
with env:
init_video_intelligence()
client = videointelligence.VideoIntelligenceServiceClient()
self.assertIsNotNone(client.credentials)
self.assertIsInstance(client.credentials, KaggleKernelCredentials)
def get_label_annotations(gcs_uri):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(gcs_uri, features=features)
# Wait until the annotate_video function call has completed.
results = operation.result(timeout=90).annotation_results[0]
label_annotations = results.segment_label_annotations
return label_annotations
def get_annotations(path_data):
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [
videointelligence.enums.Feature.LABEL_DETECTION,
videointelligence.enums.Feature.SHOT_CHANGE_DETECTION,
]
operation = video_client.annotate_video(input_content=path_data,
features=features)
return operation
def test_user_provided_credentials(self):
credentials = _make_credentials()
env = EnvironmentVarGuard()
env.set('KAGGLE_USER_SECRETS_TOKEN', 'foobar')
env.set('KAGGLE_KERNEL_INTEGRATIONS', 'CLOUDAI')
with env:
init_video_intelligence()
client = videointelligence.VideoIntelligenceServiceClient(credentials=credentials)
self.assertNotIsInstance(client.credentials, KaggleKernelCredentials)
self.assertIsNotNone(client.credentials)
def send_video(self, credentials, tmp_video):
self.logger.info('GCP call')
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = credentials
video_client = videointelligence.VideoIntelligenceServiceClient()
operation = video_client.annotate_video(
input_content=self.read_video(tmp_video),
features=[videointelligence.enums.Feature.OBJECT_TRACKING],
location_id='us-east1')
result = operation.result(timeout=900)
return result
def analyze_labels(path):
os.chdir(path)
os.chdir("../")
"""Detect labels given a file path."""
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
with io.open(path, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(features=features,
input_content=input_content)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=90)
print('\nFinished processing.')
# Process video/segment level label annotations
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
print('Video label description: {}'.format(
segment_label.entity.description))
for category_entity in segment_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
for i, segment in enumerate(segment_label.segments):
start_time = (segment.segment.start_time_offset.seconds +
segment.segment.start_time_offset.nanos / 1e9)
end_time = (segment.segment.end_time_offset.seconds +
segment.segment.end_time_offset.nanos / 1e9)
positions = '{}s to {}s'.format(start_time, end_time)
confidence = segment.confidence
print('\tSegment {}: {}'.format(i, positions))
print('\tConfidence: {}'.format(confidence))
print('\n')
#I use this code to process frame level label annotations
labels = []
frame_labels = result.annotation_results[0].frame_label_annotations
for i, frame_label in enumerate(frame_labels):
print('Frame label description: {}'.format(
frame_label.entity.description))
for category_entity in frame_label.category_entities:
print('\tLabel category description: {}'.format(
category_entity.description))
# We all know that every frame label annotation, so we just print imformation for the first frame.
frame = frame_label.frames[0]
time_offset = frame.time_offset.seconds + frame.time_offset.nanos / 1e9
print('\tFirst frame time offset: {}s'.format(time_offset))
print('\tFirst frame confidence: {}'.format(frame.confidence))
print('\n')
return labels
def run_quickstart(file_dir):
# [START video_quickstart]
i = 0
j = 0
k = 0
l = 0
errors = {}
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
for direc in os.listdir(file_dir):
print(direc)
loc=os.path.join(file_dir,direc)
os.chdir(loc)
try:
if glob.glob('*.mp4'):
file = glob.glob('*.mp4')[0]
file_name = os.path.join(file_dir,direc,file)
print("file name = " ,file_name)
with io.open(file_name, 'rb') as movie:
input_content = movie.read()
operation = video_client.annotate_video(input_content=input_content, features=features)
print('\nProcessing video for label annotations:')
result = operation.result(timeout=120)
print('\nFinished processing.')
result_json = {}
result_json = get_details(result)
with open(os.path.join(file_dir,direc,'data.json'), 'w') as outfile:
json.dump(result_json, outfile)
i = i+1
elif glob.glob('*.webm'):
l = l+1
except Exception as o:
print(o)
j = j+1
errors[loc]=str(o)
k = k + 1
print("Out of total {}, processed {} mp4 videos successfully, while {} files errored out, and left {} webm files out!".format(k,i,j, l))
with open(os.path.join(file_dir,'error/error.json'), 'w') as outfile:
json.dump(errors, outfile)
def find_labels(self):
from google.cloud import videointelligence
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(
self.gcs_uri, features=features)
result = operation.result(timeout=120)
# Single video is being processed. '''annotation_results[0]'''
self.segment_labels = result.annotation_results[0] \
.segment_label_annotations
def analyze_labels(index):
dict={}
path="gs://devfest-gjm/Vines/"+str(index)+".mp4"
video_client = videointelligence.VideoIntelligenceServiceClient()
features = [videointelligence.enums.Feature.LABEL_DETECTION]
operation = video_client.annotate_video(path, features=features)
result = operation.result(timeout=90)
segment_labels = result.annotation_results[0].segment_label_annotations
for i, segment_label in enumerate(segment_labels):
for i, segment in enumerate(segment_label.segments):
dict[str(segment_label.entity.description)]=segment.confidence
return dict
