Capturing Massive Volumes of Video Streams with AWS for Smart City Initiatives

Stream Metadata and Custom Video Processing in Hyderabad's Smart City Infrastructure

Question

As a part of the smart city initiatives, Hyderabad (GHMC), one of the largest cities in southern India is working on capturing massive volumes of video streams 24/7 captured from the large numbers of “Vivotek IB9371 - HT” cameras installed at traffic lights, parking lots, shopping malls, and just about every public venue to help solve traffic problems, help prevent crime, dispatch emergency responders, and much more.

GHMC uses AWS to host their entire infrastructure. The camera's write stream into Kinesis Video Stream securely and eventually consumed by applications for custom video processing, on-demand video playback and also consumed by AWS Rekognition for video analytics.

Along with the stream, different modes of streaming metadata are sent along with the stream.

There are 2 scenarios that need to be fulfilled. ·Requirement 1 - Affix metadata on a specific Adhoc basis to fragments in a stream, aka when smart camera detects motion in restricted areas, adds metadata [Motion = true] to the corresponding fragments that contain the motion before sending the fragments to its Kinesis Video Stream ·Requirement 2 - affix metadata to successive, consecutive fragments in a stream based on a continuing need, aka all smart cameras in the city sends the current latitude and longitude coordinates associated with all fragments it sends to its Kinesis Video Stream How can this be achieved? Select 2 options.

Answers

Explanations

Click on the arrows to vote for the correct answer

A. B. C. D. E. F.

Answer: A, D.

Option A is correct.

Requirement 1 can be fulfilled by sending Nonpersistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

Option B is incorrect.

Requirement 2 cannot be fulfilled by sending Nonpersistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

Option C is incorrect.

Requirement 1 cannot be fulfilled by sending Persistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

Option D is correct.

Requirement 2 can be fulfilled by sending Persistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

Option E is incorrect.

Requirement 1 can be fulfilled by sending Nonpersistent data while Requirement 2 can be fulfilled by sending Persistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

Option F is incorrect.

Requirement 1 cannot be fulfilled by sending Persistent data while Requirement 2 can be fulfilled by sending Persistent data.

https://docs.aws.amazon.com/kinesisvideostreams/latest/dg/how-meta.html

To fulfill Requirement 1, which is affixing metadata on a specific ad-hoc basis to fragments in a stream when the smart camera detects motion in restricted areas, adds metadata [Motion = true] to the corresponding fragments that contain the motion before sending the fragments to its Kinesis Video Stream. This can be achieved by sending Persistent data, which means the data is stored in Kinesis Video Stream and remains available even if the application stops consuming the data. With Persistent data, the metadata will remain attached to the specific fragment, and any application consuming the stream can access it. Non-persistent data, on the other hand, is stored in memory and is lost if the application stops consuming the data.

To fulfill Requirement 2, which is affixing metadata to successive, consecutive fragments in a stream based on a continuing need, such as all smart cameras in the city sending the current latitude and longitude coordinates associated with all fragments it sends to its Kinesis Video Stream. This can be achieved by sending Non-persistent data because the metadata is not attached to a specific fragment but rather to all successive fragments in the stream. Non-persistent data is a good option here because it is efficient and does not require storage space in Kinesis Video Stream. The application consuming the stream can simply look for the metadata in each successive fragment.

In summary, to fulfill Requirement 1, we need to send Persistent data, and to fulfill Requirement 2, we need to send Non-persistent data. The correct answers are:

C. Requirement 1 can be fulfilled by sending Persistent data B. Requirement 2 can be fulfilled by sending Non-persistent data.