IoT Edge 上的实时视频分析发行说明Live Video Analytics on IoT Edge release notes

本文提供以下事项的信息:This article provides you with information about:

  • 最新版本The latest releases
  • 已知问题Known issues
  • Bug 修复Bug fixes
  • 已弃用的功能Deprecated functionality

2020 年 12 月 14 日December 14, 2020

此版本是 IoT Edge 上实时视频分析的公共预览刷新版本。This release is the public preview refresh release of Live Video Analytics on IoT Edge. 发行标记为The release tag is

     mcr.microsoft.com/media/live-video-analytics:2.0.0

模块更新Module updates

  • 添加了对每个图形拓扑使用多个 HTTP 扩展处理器和 gRPC 扩展处理器的支持。Added support for using more than one HTTP extension processor and gRPC extension processor per graph topology.
  • 添加了对接收器节点的磁盘空间管理的支持。Added support for Disk space management for sink nodes.
  • MediaGraphGrpcExtension 节点现在支持 extensionConfiguration 属性,允许在单个 gRPC 服务器中使用多个 AI 模型。MediaGraphGrpcExtension node now supports extensionConfiguration property for using multiple AI models within a single gRPC server.
  • 添加了对以 Prometheus 格式收集实时视频分析模块指标的支持Added support of collecting Live Video Analytics module metrics in the Prometheus format
  • 帧速率筛选器处理器已弃用。Frame Rate Filter processor is deprecated.
    • 帧速率管理现在在图形扩展处理器节点本身中可用。Frame rate management is now available within the graph extension processor nodes itself.

2020 年 9 月 22 日September 22, 2020

此模块 2020 年 9 月刷新的此版本标记为:This release tag is for the September 2020 refresh of the module is:

mcr.microsoft.com/media/live-video-analytics:1.0.4

备注

在快速入门和教程中,部署清单使用标记 1 (live-video-analytics:1)。In the quickstarts and tutorials, the deployment manifests use a tag of 1 (live-video-analytics:1). 因此,只需重新部署此类清单,即可在“边缘”>“设备”上更新该模块。So simply redeploying such manifests should update the module on your edge > devices.

模块更新Module updates

  • 添加了对 Linux ARM64 设备的支持 - 使用手动步骤部署到此类设备。Added support for Linux ARM64 devices - use manual steps for deploying to such devices.

2020 年 8 月 19 日August 19, 2020

此模块 2020 年 8 月刷新的此版本标记为:This release tag is for the August 2020 refresh of the module is:

mcr.microsoft.com/media/live-video-analytics:1.0.3

备注

在快速入门和教程中,部署清单使用标记 1 (live-video-analytics:1)。In the quickstarts and tutorials, the deployment manifests use a tag of 1 (live-video-analytics:1). 因此,只需重新部署此类清单,即可在“边缘”>“设备”上更新该模块。So simply redeploying such manifests should update the module on your edge > devices.

模块更新Module updates

  • 现在,你可以使用 gRPC 框架在 IoT Edge 的实时视频分析和自定义扩展之间获得高数据内容传输性能。You can now get high data content transfer performance between Live Video Analytics on IoT Edge and your custom extension using gRPC framework. 请参阅此内容以开始使用。See this to get started.
  • 实时视频分析的区域部署更广泛,且只更新了云服务。Broader regional deployment of Live Video Analytics and only the cloud service has been updated.
  • 实时视频分析现已可用。Live Video Analytics is now available. 下面是所有可用区域的列表Here is the list of all available regions.
  • 快速入门的设置已更新,并支持新区域。The set up for quick starts has been updated as well with new regions support.
    • 对于已设置资源的用户,不需要进行任何操作There is no call to action for anyone who has already setup resources

Bug 修复Bug fixes

  • 在设置脚本中不再使用已弃用的 Azure 扩展Remove the use of a deprecated Azure extension in the set up script

2020 年 7 月 13 日July 13, 2020

此模块 2020 年 7 月刷新的此版本标记为:This release tag is for the July 2020 refresh of the module is:

mcr.microsoft.com/media/live-video-analytics:1.0.2

备注

在快速入门和教程中,部署清单使用标记 1 (live-video-analytics:1)。In the quickstarts and tutorials, the deployment manifests use a tag of 1 (live-video-analytics:1). 因此,只需重新部署此类清单,即可在“边缘”>“设备”上更新该模块。So simply redeploying such manifests should update the module on your edge > devices.

模块更新Module updates

  • 现在,你可以创建具有资产接收器节点以及信号门处理器节点下游的文件接收器节点的图形拓扑。You can now create graph topologies that have an asset sink node as well as a file sink node downstream of a signal gate processor node. 请参阅此内容获取示例。See this for an example.

Bug 修复Bug fixes

  • 对所需属性验证的改进Improvements to validation of desired properties

2020 年 6 月 1 日June 1, 2020

此版本是 IoT Edge 上的实时视频分析的第一个公共预览版本。This release is the first public preview release of Live Video Analytics on IoT Edge. 发行标记为The release tag is

     mcr.microsoft.com/media/live-video-analytics:1.0.0

支持的功能Supported functionalities

  • 使用所选的 AI 模块分析实时视频流,还可以选择在边缘设备上或云中录制视频Analyze live video streams using AI modules of your choice and optionally record video on the edge device or in the cloud

  • 在 IoT Edge 支持的 Linux AMD64 操作系统上使用Use on Linux AMD64 operating systems supported by IoT Edge

  • 使用 Azure 门户或 Visual Studio Code 通过 IoT 中心部署和配置模块Deploy and configure the module via the IoT Hub using Azure portal or Visual Studio Code

  • 通过以下直接方法调用来远程或本地管理图形拓扑和图形实例Manage graph topologies and graph instances remotely or locally through the following direct method calls

    • GraphTopologyGetGraphTopologyGet
    • GraphTopologySetGraphTopologySet
    • GraphTopologyDeleteGraphTopologyDelete
    • GraphTopologyListGraphTopologyList
    • GraphInstanceGetGraphInstanceGet
    • GraphInstanceSetGraphInstanceSet
    • GraphInstanceDeleteGraphInstanceDelete
    • GraphInstanceListGraphInstanceList

后续步骤Next steps

概述Overview