Video Annotation Tutorial

UVAP video annotation is processing pre-recorded videos by annotating them with the visual results of different features of UVAP.

Prerequisites

Before continuing with the tutorial, ensure the following:

  • UVAP is installed as instructed in Setting Up UVAP
  • No UVAP-related Docker containers are running

Preparing Source Video and Containers

  1. Create the source directory for videos:

    $ mkdir -p "${UVAP_HOME}"/videos

  2. Copy the source video to the created directory:

    $ cp -i <CURRENT_VIDEO_PATH>/testvideo.mkv \
  "${UVAP_HOME}"/videos/testvideo.mkv

    Note: By using -i for interactive, the user is asked if the file is to be replaced (if already exists).

  3. Check video resolution and frame rate:

    $ ffprobe "${UVAP_HOME}"/videos/testvideo.mkv

Note: Important part of the output of ffprobe:

[...]
    Stream #0:0(eng): Video: h264 (Main), yuvj420p(pc, bt709, progressive), 1280x720 [SAR 1:1 DAR 16:9], 6 fps, 6 tbr, 1k tbn, 2k tbc (default)
[...]

Where:

  • Resolution is 1280 × 720 (width × height)
  • Frame rate is 6 fps

  1. Stop, remove, and restart Zookeeper and Kafka containers with the following command:

    Attention! This process removes all topics from Kafka.

    $ "${UVAP_HOME}"/scripts/restart_empty_kafka.sh

Demo Usage for Video Annotation

Available Arguments

List available arguments of video annotation command with the following command:

$ "${UVAP_HOME}"/scripts/video_annotation.sh --help

Expected output:

Usage: ./video_annotation.sh [OPTION]...
Options (those, that do not have a default value, are mandatory options):
    --demo-name                                     <head_detection|head_pose|demography|tracker|detection_filter|skeleton|reidentification>
    --video-dir                                     directory path of the input and output videos - default: /home/ultinous/uvap/scripts/../videos
    --video-file-name                               basename of the input video file inside the --video-dir directory
    --fps-number                                    the FPS number of the input video
    --width-number                                  the width in pixels of the input video
    --height-number                                 the height in pixels of the input video
    --demo-applications-dir                         directory path of demo applications scripts - default: /home/ultinous/uvap/scripts/../demo_applications
    --templates-dir                                 directory path of configuration templates - default: /home/ultinous/uvap/scripts/../templates
    --config-ac-dir                                 directory path of configuration files - will be created if not existent - default: /home/ultinous/uvap/scripts/../config
    --models-dir                                    directory path of AI models - default: /home/ultinous/uvap/scripts/../models
    --license-data-file                             data file of your UVAP license - default: /home/ultinous/uvap/scripts/../license/license.txt
    --license-key-file                              key file of your UVAP license - default: /home/ultinous/uvap/scripts/../license/license.key
    --uvap-mgr-docker-image-name                    tag of docker image to use - default: will be determined by git tags
    --uvap-demo-applications-docker-image-name      tag of docker image to use - default: will be determined by git tags
    --uvap-kafka-tracker-docker-image-name          tag of docker image to use - default: will be determined by git tags
    --uvap-kafka-detection-filter-docker-image-name tag of docker image to use - default: will be determined by git tags
    --uvap-kafka-reid-docker-image-name             tag of docker image to use - default: will be determined by git tags
    --display                                       option to display the video during the annotation
    --verbose

Note: Options that do not have a default value are mandatory.

Annotating the Video

Use the video writer script to create the output video into "${UVAP_HOME}"/videos with [IMAGE_TOPIC_NAME].avi name, overwriting the existing file:

$ "${UVAP_HOME}"/scripts/video_annotation.sh \
  --video-file-name [INPUT_VIDEO] \
  --fps-number [FPS] \
  --width-number [WIDTH] \
  --height-number [HEIGHT] \
  --demo-name [DEMO_NAME]

Parameters:

  • [INPUT_VIDEO]: name of the input video. Replace parameter with the name of the existing video, following the example above: testvideo.mkv.
  • [FPS], [WIDTH], and [HEIGHT]: parameters of output video. Replace with the parameters of the original video, see Preparing Source Video and Containers.
  • [DEMO_NAME]: name of the demo. Replace (including the brackets) with a name of the demo (string). Valid values of [DEMO_NAME] are the following:
    • head_detection
    • head_pose
    • demography
    • face_mask
    • tracker
    • detection_filter
    • skeleton
    • reidentification

Expected console output:

Starting to configure...
Finished configuring.
Starting the core analysis...
Finished core analysing.
Starting to annotate...
Finished annotating.
Starting to write video...
Finished writing video.

Default path of result video:

"${UVAP_HOME}"/videos/[DEMO_MODE].cam.0.[DEMO_NAME].Image.jpg.avi
← Example Analysis Results