媒体服务平台上的媒体分析Media Analytics on the Media Services platform

停用计划Retirement plans

重要

一些媒体处理器即将停用。Some media processors are being retired. 有关停用日期和详细信息,请参阅旧组件主题。For the retirement dates and more information, see the legacy components topic.

概述Overview

越来越多的组织将视频用作首选媒体来培训员工、与客户沟通以及规定业务职能。More organizations are using video as the preferred medium to train their employees, engage their customers, and document business functions. 云计算提供了一种方法来存储、流式传输和访问这些大型媒体文件。Cloud computing provides a way to store, stream, and access these large media files. 但随着公司视频内容库的扩大,它需要在从内容中提取见解时具有同样有效的方式。But as a company's library of video content grows, it needs an equally effective means of extracting insights from the content.

为了满足这一不断增长的需求,Azure 媒体服务提供了 Azure 媒体分析。To address this growing need, Azure Media Services offers Azure Media Analytics. 媒体分析是语音和视觉组件的集合,便于组织和企业从视频文件中汲取可以实施的见解。Media Analytics is a collection of speech and vision components that makes it easier for organizations and enterprises to derive actionable insights from their video files. 媒体分析通过使用核心媒体服务平台组件构建而成,一开始就能进行大规模的媒体处理。Built by using the core Media Services platform components, Media Analytics can handle media processing at scale on day one.

借助媒体分析,开发人员可以快速将高级视频功能引入应用程序中。With Media Analytics, developers can quickly bring advanced video functionality into applications. 它提供企业环境,可满足各种规模的大型组织所需的合规性、安全性和全球市场宣传需求。It provides enterprise environments with the full scale, compliance, security, and global reach required by large organizations.

下图显示媒体分析和媒体服务平台的其他主要部分。The following diagram shows Media Analytics and other major parts of the Media Services platform.

VoD 工作流

媒体分析媒体处理器会生成 MP4 文件或 JSON 文件。Media Analytics media processors produce MP4 files or JSON files. 如果媒体处理器生成了 MP4 文件,可以采用渐进方式下载该文件。If a media processor produces an MP4 file, you can progressively download the file. 如果媒体处理器生成了 JSON 文件,可以从 Azure Blob 存储下载该文件。If a media processor produces a JSON file, you can download the file from Azure Blob storage.

媒体分析服务Media Analytics services

索引器Indexer

借助 Azure Media Indexer 可使内容可搜索并生成隐藏式字幕跟踪。With Azure Media Indexer, you can make content searchable and generate closed-captioning tracks. 有关详细信息和示例,请参阅使用 Azure Media Indexer 为媒体文件编制索引For detailed information and examples, see Indexing Media Files with Azure Media Indexer.

动作检测器Motion Detector

可以使用移动探测器来检测背景平稳的视频中的动作。You can use Motion Detector to detect motion in a video with stationary backgrounds. 这样,便可以检查监控摄像机所检测到的动作事件是否存在误报。This makes it possible to check for false positives on motion events detected by surveillance cameras. 有关详细信息和示例,请参阅 Azure 媒体分析的动作检测For detailed information and examples, see Motion detection for Azure Media Analytics.

面部检测器Face Detector

使用面部检测器可以检测人脸及情绪,包括快乐、悲伤和惊讶。By using Face Detector, you can detect people’s faces and their emotions, including happiness, sadness, and surprise. 此服务具有多个有用行业应用程序(稍后将进行介绍),包括聚合与分析参与事件的人员的反应。This has several useful industry applications, described later, including aggregating and analyzing reactions of people attending an event. 有关详细信息和示例,请参阅 Azure 媒体分析的人脸和情绪检测For detailed information and examples, see Face and emotion detection for Azure Media Analytics.

视频摘要Video summarization

视频摘要可通过自动选择来自源视频的有趣片段帮助你创建长视频的摘要。Video summarization can help you create summaries of long videos by automatically selecting interesting snippets from the source video. 如果想要提供有关长视频内容的快速概述,此功能会很有用。This ability is useful when you want to provide a quick overview of what to expect in a long video. 有关详细信息和示例,请参阅使用 Azure Media Video Thumbnails 创建视频摘要For detailed information and examples, see Use Azure Media Video Thumbnails to create video summarization.

光学字符识别Optical character recognition

借助 Azure 媒体 OCR(光学字符识别),可以将视频文件中的文本内容转换成可编辑、可搜索的数字文本。With Azure Media OCR (optical character recognition), you can convert text content in video files into editable, searchable digital text. 可以从媒体的视频信号中自动提取有意义的元数据。You can then automate the extraction of meaningful metadata from the video signal of your media.

可缩放的面部修订Scalable face redaction

Azure 媒体修订器是一种媒体分析媒体处理器,可用于在云中进行可缩放的面部修订。Azure Media Redactor is a Media Analytics media processor that offers scalable face redaction in the cloud. 使用面部修订,可对视频进行修改,使所选个人的面部模糊显示。By using face redaction, you can modify your video to blur faces of selected individuals. 用户可能想要在新闻媒体中或涉及公共安全时使用面部修订服务。You might want to use the face redaction service in news media or when public safety is involved. 对于时长仅几分钟但包含多张面孔的镜头,进行手动面部修订可能需要几个小时,但使用此服务仅需几个简单步骤即可完成面部修订。A few minutes of footage that contains multiple faces can take hours to redact manually, but with this service, face redaction takes just a few simple steps. 有关详细信息,请参阅使用 Azure 媒体分析进行面部修订文章。For more information, see the Redact faces with Azure Media Analytics article.

内容审核Content Moderation

借助 Azure 内容审查器,可以对视频使用机器辅助审查。Azure Content Moderator enables you to use machine-assisted moderation for your videos. 例如,你可能希望检测视频中可能存在的成人和猥亵内容,以及复查由人工审查团队标记出的内容。For example, you might want to detect possible adult and racy content in videos and review the flagged content by your human moderation teams. 手动审查不良内容的视频是一项耗时且成本高昂的任务。Manually moderating videos for undesirable content is a time consuming and expensive task. 使用此服务及关联的检查工具,就可将机器辅助审查与“人在回路中”功能相结合,以经济高效的方式获得最佳结果。With this service and associated review tools, you combine machine-assisted moderation with human-in-the-loop capabilities for best results efficiently and cost-effectively. 若要了解详细信息,请参阅使用 Azure 内容审查器处理视频一文。To learn more, see the Process your videos with Azure Content Moderator article.

常见方案Common scenarios

媒体分析可帮助组织和企业从视频搜集新的见解,并更有效地管理大量视频内容。Media Analytics can help organizations and enterprises glean new insights from video and more effectively manage large volumes of video content. 下面介绍几种方案:Here are several scenarios:

  • 呼叫中心Call centers. 即使出现了社交媒体,客户呼叫中心仍能帮助解决大量的客户服务事务。Even with the advent of social media, customer call centers still facilitate a large percentage of customer-service transactions. 此音频数据中编码了大量客户信息,分析这些信息可实现更高的客户满意度。Encoded in this audio data is a large amount of customer information that can be analyzed to achieve higher customer satisfaction. 通过使用媒体索引器,组织可以提取文本并生成搜索索引和仪表板。By using Media Indexer, organizations can extract text and build search indexes and dashboards. 然后,他们可以提取围绕常见的投诉、投诉原因以及其他相关数据的情报。Then they can extract intelligence around common complaints, sources of complaints, and other relevant data.
  • 用户生成的内容仲裁User-generated content moderation. 从新闻媒体分支机构到公安部门,许多组织都设立了对外公开的门户,用于在其中接受用户生成的媒体,例如视频和图像。From news media outlets to police departments, many organizations have public-facing portals that accept user-generated media such as videos and images. 内容的数量可能因意外事件而激增。The volume of content can spike due to unexpected events. 在这些情况下,仅凭人力几乎不可能有效地审查内容的适宜性。In these scenarios, it is difficult to conduct effective manual reviews of content for appropriateness. 客户可以依赖于内容仲裁服务,将工作重点放在适当的内容上。Customers can rely on the content-moderation service to focus on content that is appropriate.

媒体分析媒体处理器Media Analytics media processors

本节列出媒体分析媒体处理器,并说明如何使用 .NET 或 REST 来获取媒体处理器 (MP) 对象。This section lists the Media Analytics media processors and shows how to use .NET or REST to get a media processor (MP) object.

MP 名称MP names

  • Azure 媒体索引器Azure Media Indexer
  • Azure 媒体面部检测器Azure Media Face Detector
  • Azure Media Motion DetectorAzure Media Motion Detector
  • Azure 媒体视频缩略图Azure Media Video Thumbnails
  • Azure 媒体 OCRAzure Media OCR
  • Azure 媒体内容审查器Azure Media Content Moderator

.NET.NET

以下函数采用其中一个指定的 MP 名称,并返回 MP 对象。The following function takes one of the specified MP names and returns an MP object.

static IMediaProcessor GetLatestMediaProcessorByName(string mediaProcessorName)
{
    var processor = _context.MediaProcessors
        .Where(p => p.Name == mediaProcessorName)
        .ToList()
        .OrderBy(p => new Version(p.Version))
        .LastOrDefault();

    if (processor == null)
        throw new ArgumentException(string.Format("Unknown media processor",
                                                  mediaProcessorName));

    return processor;
}

RESTREST

请求:Request:

GET https://media.chinacloudapi.cn/api/MediaProcessors()?$filter=Name%20eq%20'Azure%20Media%20OCR' HTTP/1.1
DataServiceVersion: 1.0;NetFx
MaxDataServiceVersion: 3.0;NetFx
Accept: application/json
Accept-Charset: UTF-8
User-Agent: Microsoft ADO.NET Data Services
Authorization: Bearer <token>
x-ms-version: 2.19
Host: media.chinacloudapi.cn

响应:Response:

. . .

{  
    "odata.metadata":"https://media.chinacloudapi.cn/api/$metadata#MediaProcessors",
    "value":[  
        {  
            "Id":"nb:mpid:UUID:074c3899-d9fb-448f-9ae1-4ebcbe633056",
            "Description":"Azure Media OCR",
            "Name":"Azure Media OCR",
            "Sku":"",
            "Vendor":"Microsoft",
            "Version":"1.1"
        }
    ]
}

演示Demos

请参阅 Azure 媒体分析演示See Azure Media Analytics demos.

请参阅媒体服务分析公告See Media Services Analytics announcement.

后续步骤Next steps

查看媒体服务学习路径。Review Media Services learning paths.

媒体服务 v3(最新版本)Media Services v3 (latest)

查看最新版本的 Azure 媒体服务!Check out the latest version of Azure Media Services!

媒体服务 v2(旧版)Media Services v2 (legacy)