Source monitoring for AWS Elemental MediaLive via Amazon Rekognition
Media Blog
This article discusses a solution for source monitoring for AWS Elemental MediaLive using Amazon Rekognition. It allows detecting impaired video input sources, like black video, color bars, or frozen frames, which would otherwise go undetected.
Specifically, the article covers:
- Leveraging MediaLive's input source thumbnails and passing them to Amazon Rekognition for image properties analysis
- Using Rekognition's response data (brightness, sharpness, contrast) to detect anomalies like black frames or frozen input
- A basic solution using AWS Lambda triggered by CloudWatch Scheduler to poll thumbnails periodically
- An advanced solution using AWS Fargate for lower latency detection, handling multiple channels more effectively
- Configuring CloudWatch alarms based on the detected anomalies to alert operators or take automated actions
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Feb 22
2024
2024
Analyze content using AWS Elemental MediaLive thumbnails with Amazon Rekognition
Feb 15
2024
2024
How to modify an incoming audio source using AWS Elemental MediaLive
Jun 19
2024
2024
AWS Elemental MediaConnect adds source stream monitoring
Apr 5
2024
2024
How to perform AWS Elemental MediaConvert job compliance checks using Amazon CloudWatch
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.