aws-samples / amazon-rekognition-video-analyzer

A working prototype for capturing frames off of a live MJPEG video stream, identifying objects in near real-time using deep learning, and triggering actions based on an objects watch list.
Other
367 stars 157 forks source link

Video takes infinite time to analyze #47

Open zebpaypeswani opened 5 years ago

zebpaypeswani commented 5 years ago

EDIT it started working now. I want to know is there any limit on how many video we can process. if yes how can i increase this limit?

Hi some time before it was working fine and I was able to analyze videos with in minute or so but suddenly my video is not getting analyze. It just goes on forever. I am trying so search faces in a videos. faces are indexed. they are about 300 face indexed

I am analyzing 3553477.mp4 (only 10 second video)

here is the output

Faces indexed:
  Face ID: a1c8847f-7e80-4dd7-8b51-63c0d22ab765
  Location: {'Width': 0.7492865920066833, 'Height': 0.8562403321266174, 'Left': 0.1664663851261139, 'Top': 0.032046884298324585}
started video analyzing 
Start Job Id: 952741e2adaf526ca8fd270912d2cf578d24b69039edcac05dcc507b7b2193fb
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................
....................

It goes on and does not stop

here is the code

import boto3
import json
import sys

class VideoDetect:
    rek = boto3.client('rekognition')
    queueUrl = 'https://sqs.us-west-2.amazonaws.com/655136467581/Rekoginition'
    roleArn = 'arn:aws:iam::655136467581:role/face_matching'
    topicArn = 'arn:aws:sns:us-west-2:655136467581:image-rekoginition-sns'
    bucket = 'face-reckognition-video-1'
    video = ''

    def __init__(self, video):
        self.video = video

    def main(self):

        jobFound = False
        sqs = boto3.client('sqs')

        # =====================================
        response = self.rek.start_face_search(Video={'S3Object': {'Bucket': self.bucket, 'Name': self.video}},
                                              CollectionId='FaceCollection',
                                              NotificationChannel={'RoleArn': self.roleArn,
                                                                   'SNSTopicArn': self.topicArn})

        # =====================================
        print('Start Job Id: ' + response['JobId'])
        dotLine = 0

        faces = set()
        while not jobFound:
            sqsResponse = sqs.receive_message(QueueUrl=self.queueUrl, MessageAttributeNames=['ALL'],
                                              MaxNumberOfMessages=10)

            if sqsResponse:

                if 'Messages' not in sqsResponse:
                    if dotLine < 20:
                        print('.', end='')
                        dotLine = dotLine + 1
                    else:
                        print()
                        dotLine = 0
                    sys.stdout.flush()
                    continue

                for message in sqsResponse['Messages']:
                    notification = json.loads(message['Body'])
                    rekMessage = json.loads(notification['Message'])
                    print(rekMessage['JobId'])
                    print(rekMessage['Status'])
                    if str(rekMessage['JobId']) == response['JobId']:
                        print('Matching Job Found:' + rekMessage['JobId'])
                        jobFound = True
                        # =============================================
                        f = self.GetResultsFaceSearchCollection(rekMessage['JobId'])

                        faces.update(f)

                        # =============================================

                        sqs.delete_message(QueueUrl=self.queueUrl,
                                           ReceiptHandle=message['ReceiptHandle'])
                    else:
                        print("Job didn't match:" +
                              str(rekMessage['JobId']) + ' : ' + str(response['JobId']))
                    # Delete the unknown message. Consider sending to dead letter queue
                    sqs.delete_message(QueueUrl=self.queueUrl,
                                       ReceiptHandle=message['ReceiptHandle'])

        print('done')

        return faces

    def GetResultsFaceSearchCollection(self, jobId):
        maxResults = 10
        paginationToken = ''
        faces = []
        finished = False

        while not finished:
            response = self.rek.get_face_search(JobId=jobId,
                                                MaxResults=maxResults,
                                                NextToken=paginationToken)
            print(response)
            print(response['VideoMetadata']['Codec'])
            print(str(response['VideoMetadata']['DurationMillis']))
            print(response['VideoMetadata']['Format'])
            print(response['VideoMetadata']['FrameRate'])

            for personMatch in response['Persons']:

                print('Person Index: ' + str(personMatch['Person']['Index']))
                print('Timestamp: ' + str(personMatch['Timestamp']))

                if 'FaceMatches' in personMatch:
                    for faceMatch in personMatch['FaceMatches']:
                        print('Face ID: ' + faceMatch['Face']['FaceId'])
                        faces.append(faceMatch['Face']['FaceId'])
                        print('Similarity: ' + str(faceMatch['Similarity']))
                print()
            if 'NextToken' in response:
                paginationToken = response['NextToken']
            else:
                finished = True
            print()

        return faces
shezgone commented 5 years ago

I have the same problem.

kyle861861 commented 4 years ago

i have too