Video Curation Sample Reference Implementation

Published: 12/02/2021  

Last Updated: 01/13/2022

Overview

Implements a reference pipeline using libraries for content analysis of video files. Includes database ingestion, content search and visualization as a foundation.

Select Configure & Download to download the sample and the software listed below.   

Configure & Download

Image of a screen with Video Curation Concept Sample at the top. The screen shows multiple videos with an enlarged video of people sitting around an office table.

  • Time to Complete: 20 minutes
  • Programming Language: Python*, JavaScript*, Shell*, C++
  • Available Software: 
    • ​​​GStreamer
    • OpenCV
    • NGINX
    • Kafka
    • Zookeeper 
    • Intel® Distribution of OpenVINO™ toolkit

Target System Requirements 

  • Intel® Xeon® platform, 64 GB RAM or higher 
  • Recommended OS: Ubuntu* 18.04 / CentOS* 7 
  • Disk Space needed: 3.5 GB (Source: 1 GB, Docker* Images: 2.5 GB)

How It Works

This sample implements libraries of video files content analysis, database ingestion, content search and visualization:

  • Ingest: Analyze video content and ingest the data into the VDMS.
  • Visual Data Management System (VDMS): Store metadata efficiently in a graph-based database.
  • Visualization: Visualize content search based on video metadata.

Software Stacks

The sample is powered by the following Open Visual Cloud software stacks:

  • Media Analytics: The GStreamer-based media analytics stack is used for object, face and emotion detection. The software stack is optimized for Intel® Xeon® Scalable processors.
  • NGINX Web Service: The NGINX/FFmpeg-based web serving stack is used to store and segment video content and serve web services. The software stack is optimized for Intel® Xeon® Scalable processors.
A diagram with a heading of Kubernetes or Docker Swarm and arrows going from "Video Storage Service" to "Segmented Video" and to "Video." Video points to Ingest Pool, which points to Analytics, which points to VDMS. Off to the side is a block with "Scheduling" as a header. "Browser" is outside of the main block with an arrow pointing to Web Service, which has a double arrow and also points to VDMS.
Figure 1: Architecture Diagram

 


Get Started

Prerequisites

Follow the steps on GitHub to install the prerequisites.

Install the Sample

Select Configure & Download to download the sample.   

Configure & Download 

Build the Sample  

Follow the steps on GitHub to build and run the sample.

Learn More

To continue learning, see the following guides and software resources: 

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.