Amazon Kinesis Video Streams
Amazon Kinesis Video Streams - is a service that streamlines the secure transfer of video from connected devices to AWS for analytics, machine learning (ML), playback, and other processing needs. It takes care of automatically provisioning and scaling the required infrastructure to handle streaming video data from millions of devices. The service ensures durable storage, encryption, and indexing of video data within your streams, offering convenient access through user-friendly APIs. Kinesis Video Streams enables both live and on-demand video playback, allowing the development of applications that leverage computer vision and video analytics. Integration with Amazon Rekognition Video and libraries for popular ML frameworks like Apache MxNet, TensorFlow, and OpenCV further enhances its capabilities. Additionally, Kinesis Video Streams supports WebRTC, an open-source project that facilitates real-time media streaming and interaction among web browsers, mobile applications, and connected devices through simple APIs. This enables applications such as video chat and peer-to-peer media streaming.
Key definitions for Amazon Kinesis Video Streams:
-
SDKs to securely stream data from devices
Service provides SDKs in C++ and Java. SDK also can be built from sources or used in Docker images available for Ubuntu, MacOS, and Raspberry Pi.
-
Live and on-demand video playback with HTTP Live Streaming (HLS)
Amazon Kinesis Video Streams provides the capability to stream live video content in real-time and also allows for the playback of recorded video on demand. HLS is an adaptive streaming communications protocol that enables the efficient delivery of media content over the web. With HLS support, Kinesis Video Streams can deliver a seamless streaming experience that adjusts to the viewer's internet speed and device capabilities, minimizing buffering and ensuring high-quality playback.
-
Built-in integration with Amazon Rekognition Video
Offers direct integration with Amazon Rekognition Video, an advanced video analysis service. This allows users to add machine learning capabilities to their video streams effortlessly. With this integration, users can detect activities, recognize faces, and identify objects within video streams in real time or from stored video content, enabling the creation of intelligent video analytics applications.
-
Built-in image extraction functionality
The service includes functionality that enables users to extract still images from their video streams. This can be used for various applications such as generating thumbnails, performing image analysis, or capturing significant events. The image extraction is performed without affecting the video stream, ensuring that the live streaming or recorded video is not interrupted.
-
Be notified when the media is ready for consumption
Service uses integration with SNS to provide notification about video readiness.
-
Real-time APIs
Built in support of real time interaction with video streams.
-
Support for two-way low-latency video streaming based on WebRTC
Supports WebRTC, an open-source project that provides web browsers and mobile applications with real-time communication capabilities. This allows for two-way, low-latency video streaming between clients and servers. This feature is crucial for applications that require interactive video communications, such as video conferencing or telemedicine, where minimal delay is essential.
-
Durable storage
Service uses AWS S3 for storing video.
-
Pay per use
Service charges are based on data usage, without upfront costs or minimum fees. Idle video streams are not charged, providing cost-effective video streaming.
-
Automatic data encryption in transit and at rest
To ensure the security of video data, Kinesis Video Streams automatically encrypts the video data both in transit (as it is being uploaded to and downloaded from the cloud) and at rest (while it is stored on the cloud). This encryption process is managed transparently by the service, which helps in safeguarding data against unauthorized access and ensuring compliance with various privacy standards and regulations.
-
Access control with AWS IAM
Access to video streams can be controlled using AWS Identity and Access Management (IAM), which provides fine-grained access control over AWS resources. Users can define policies to grant or deny access to their Kinesis Video Streams resources, ensuring that only authorized personnel or applications can publish or consume video streams.
-
Time-encoded data
Kinesis Video Streams records and stores video streams along with associated metadata and other time-coded data. This time-encoded data can include timestamps, sensor readings, or other metadata that correlates with the video, enabling precise synchronization and making it easier to search and analyze specific segments of the video stream in relation to the collected data.
-
Fully managed
Service is fully managed, so there are no servers to manage.
Service integrates with:
Usage use cases
-
Smart home.
Amazon Kinesis Video Streams simplifies the process of streaming live video and audio from camera-equipped home devices, such as doorbells, baby monitors, webcams, and home surveillance systems, to AWS. These streams can be utilized to create a wide range of smart home applications, including video playback, intelligent lighting, climate control systems, and security monitoring. With the WebRTC capabilities offered by the service, two-way, real-time media streaming and interaction become possible. This opens up possibilities such as communicating with the person at your doorbell or remotely controlling camera-enabled devices like robot vacuums from your mobile phone.
-
Smart city.
In numerous cities, a substantial quantity of cameras have been deployed across traffic lights, parking lots, shopping malls, and various public locations, continuously capturing video footage. Amazon Kinesis Video Streams provides a secure and cost-effective solution for efficiently ingesting, storing, and analyzing large volumes of video data. By leveraging the service, cities can address traffic issues, enhance crime prevention measures, facilitate the dispatching of emergency responders, and achieve a multitude of other valuable applications and insights.
-
Industrial automation.
Amazon Kinesis Video Streams enables the collection of diverse time-encoded data, including RADAR and LIDAR signals, temperature profiles, and depth data from industrial equipment. This data can be seamlessly analyzed using popular machine learning frameworks like Apache MxNet, TensorFlow, and OpenCV, allowing for industrial automation applications such as predictive maintenance. For instance, you can predict the lifespan of a gasket or valve and proactively schedule part replacements, minimizing downtime and defects in manufacturing lines. The service empowers businesses to optimize their operations and enhance efficiency through intelligent data-driven insights.
FAQ for Amazon Kinesis Video Streams
-
Does Amazon Kinesis Video Streams support automatic data encryption?
Yes it does. Service supports automatic data encryption in transit and at rest. -
Does Amazon Kinesis Video Streams support image extraction from video streams?
Yes it does. Service provides a built-in method for extracting images for video streams. -
How is it possible to get access to server instances used by Amazon Kinesis Video Streams?
Service is fully serverless and removes complexity of server management from clients. Video streams automatically scales in and out based on created load.