Smart City Sample Versions Save

• Upgraded to docker images v21.3.
• Supported additional camera protocols: RTMP and GB28181.
• Bug fixes.

• Upgraded the v20.10 Open Visual Cloud software stack, with Intel(R) OpenVINO(TM) Toolkit v2021.1.
• Rearchitected the cloud/edge services to use cloud/edge gateways.
• Support realtime camera preview utilizing the WebRTC toolkit.

• Upgraded to v20.7 Open Visual Cloud software stacks, supporting the latest 2020.4 Intel(R) OpenVino(TM) Toolkit features.
• Supported dynamically and reliably starting/stopping multiple offices.
• Ran all containers as non-root users, (with the exception of the VCAC-A platform analytics containers.)
• Improved IP camera scanning and discovery.
• Added cloud storage UI and recording review.
• Fixed bugs in smart uploading.

• Updated to OpenVINO v2020R2 and models.
• Updated to gst-video-analytics v1.0.0.
• Updated to video-analytics-serving 0.3-preview.
• Fixed a few bugs in the Kubernetes deployment scripts.
• New NETWORK build option to specify the model network preference.

The Smart City release v20.1 includes the following stable features: (1) Smart Stadium features including people-counting, queue-counting and crowd-counting. Configure the stadium mode with cmake -DSCENARIO=stadium ..
(2) VCAC-A support for the traffic scenario for both docker-swarm and Kubernetes. Stadium supports to be followed.
(3) Reduced build time by fully integrating latest openvisualcloud v20.1 docker images.

This minor version release updated the Kubernetes deployment scripts for the multi-office case.

This release targets Smart City Traffic Analysis, including features as follows:

• Camera Discovery: Auto-discover cameras based on the ONVIF protocol. Auto-match the cameras with analytics instances.
• Traffic Analysis: Detect vehicles, bicycles and pedestrians for traffic analysis and city planning.
• Inline Recording: Record camera feeds into segments for review at a later time.
• Smart Upload: Selectively upload recordings to the cloud for post-processing or archival based on analytics results.
• Deployment Support: Support docker compose (single-node development), docker swarm (single or multi-node deployment), and Kubernetes (multi-node deployment.)
• Platform Support: Support the Intel(R) Xeon(R) and Intel VCAC-A (accelerator) platforms.
• New UI: Reworked UI hopefully you will enjoy.

Please see README.md at root or under each folder for more details about the features, usages and setup.

The release features the following improvements: -- UI improvements: Drag & drop previews and stats charts onto map. -- Divide the services into office services, cloud services and camera services, each of which can be on a separate network, for scalability. -- VA service points to v1.1.

Release Notes:

The E2E sample implements smart city use cases based video analytics. In this release, the following services are implemented.

The Video Analytic Service The Web Cloud service The Web Local service The Simulated Camera service The Camera Discovery service The Where Indexing service The Storage Cleanup service The Health Check service