↑ Return to Software + DATA

CAR (Complex Activity Recognition) module

CAR(Complex Activity Recognition), written in C++ , is designed for analyzing video content . CAR is able to recognize events such as ‘falling’, ‘walking’ of a person. CAR divides the work-flow of a video processing into several separated modules, such as acquisition, segmentation, up to activity recognition. Each module has a specific interface, and different plugins (corresponding to algorithms) can be implemented for a same module. We can easily build new analyzing systems thanks to this set of plugins. The order we can use those plugins and their parameters can be changed at run time and the result visualized on a dedicated GUI. This platform has many more advantages such as easy serialization to save and replay a scene, portability to Mac, Windows or Linux, and easy deployment to quickly setup an experimentation anywhere. CAR takes different kinds of input: RGB camera, depth sensor for online processing; or image/video files for offline processing.

This generic architecture is designed to facilitate:

  1. integration of new algorithms into CAR;

  2. Sharing of the algorithms among the Stars team. Currently, 15 plugins are available, covering the whole processing chain. Some plugins use the OpenCV library.

The plugins cover the following research topics:

  • algorithms : 2D/3D mobile object detection, camera calibration, reference image updating, 2D/3D mobile object classification, sensor fusion, 3D mobile object classification into physical objects (individual, group of individuals, crowd), posture detection, frame to frame tracking, long-term tracking of individuals, groups of people or crowd, global tacking, basic event detection (for example entering a zone, falling…), human behaviour recognition (for example vandalism, fighting,…) and event fusion; 2D & 3D visualisation of simulated temporal scenes and of real scene interpretation results; evaluation of object detection, tracking and event recognition; image acquisition (RGB and RGBD cameras) and storage; video processing supervision; data mining and knowledge discovery; image/video indexation and retrieval.

  • languages : scenario description, empty 3D scene model description, video processing and understanding operator description;

  • knowledge bases : scenario models and empty 3D scene models;

  • learning techniques for event detection and human behaviour recognition;

Installation of CAR from scratch:

Running CAR :

Tools / tutorials :

FAQ :

Start CAR automatically and remotely

The following document is a guide to start CAR component automatically. You need to make start and stop.sh and modify some configuration files in your Desktop environment. To simplify the management, you should create a scripts folder in your home folder (e.g. mkdir ~/demcare/scripts). Summary : Creating start and stop scripts To start CAR with …

View page »

Deploment of CAR in Lab

Requirements : Dem@care system installed on Windows 7 x64 CAR component installed on Fedora A network connection between CAR component and Dem@care system Lab deployment: Download the config folders for the room and corridor scene ConfigFile_Lab to the location of your choice e.g. /home/username/demacare/ You may wish to use these options avoid recording RGB videos but …

View page »

Advised hardware for CAR

RGB-D Sensor Brand : ASUS Designation : Xtion PRO LIVE Model : 90IW0122-B01UA Technical spécifications : Interface / power : USB2.0 Sensor : RGB & Depth Audio : built-in microphones*2 where to buy : http://www.ldlc.com/fiche/PB00117374.html Computer Brand : Dell Designation : OptiPlex 9020 Model : OptiPlex 9020 Ultra Small Form Factor Technical spécifications : Processor : Intel Core i5 4570 Memory : 8GB (2 x 4GB DDR3) Storage : …

View page »

CAR – Scene context creation tool

Summary : Decompress depth and colour image Create the file sup_context.xml For previous versions of tools Preparation : Kreat-tool will help create the scene context in a xml file. You will draw important zone, used for event modelling. WARNING: The calibration file calibration.yml has been changed on the 10th of Septembre 2013 due to two swapped …

View page »

CAR – Testing the CAR component

Content: 1 – Data structure 2 – Testing the video data recorder 3 – Testing the video data reader 4 – Running the CAR Dem@Care application 5 – Testing on an existing video 1 – Data structure The CAR component provides one application called ‘sup-3DSensor_writer’ to continuously write compressed data into local hard drive. The …

View page »

CAR installer script

Summary : About the script CAR-installer.sh Install CAR from scratch 1 – About the script : 1.1 – Script usage Get help by using the -h or –help parameters. ./car-installer.sh [parameter] 1.2 – Script function By defaut the script install CAR from scratch, so you can start the script without parameter if you simply want …

View page »

Deployment of CAR in Nursing Home

Requirements : Dem@care system installed on Windows 7 x64 CAR component installed on Fedora A network connection between CAR component and Dem@care system Nursing Home deployment: Download the config folder ConfigFile_NursingHome to the location of your choice e.g. /home/username/demacare/ You may wish to avoid recording RGB videos but only depth videos for privacy reasons. To …

View page »

How to view and backup your postgresql database

You can view and backup your postgresql database using A – the pgadmin3 interface B – a command line ———————————————————————————————————————————– A – How to view and backup your postgresql database using the pgadmin3 interface If pgadmin3 is not install, install it as follows $ su # yum install pgadmin3 # exit # pgadmin3 Before launching …

View page »

Install CAR Component on Fedora 19 (in progress)

Summary : Install ASUS Sensor and OpenNI Install PostgreSQL Database Install SUP packages Test and use the CAR component Requirements : You have a machine with Fedora 19 64bits installed (get fedora 19 64bits). A network access on Internet is necessary. A user with administrative rights (with sudo or root) is necessary. You need wget …

View page »

Install Fedora system

Summary : Get your Live USB maker Create a bootable Live USB drives with UNetbootin Installation of Fedora Required : You have a USB key (at least 1 GB for a Live image of Fedora). A network access to Internet is necessary to get the ISO. A PC to make your Fedora Live USB. A …

View page »

Installation of CAR on Linux Fedora 17 from scratch:

Summary : Install PostgreSQL Database Install CAR package for the first time Requirements : You have a machine with Fedora 17 64bits installed. (get fedora 17 64bits) A network access on Internet is necessary. A user with administrative rights (with sudo or root) is necessary. You need wget and unzip programs (yum install wget unzip). …

View page »

Installation of CAR on Linux Fedora 19 from scratch

Summary : Install ASUS Sensor and OpenNI Install PostgreSQL Database Install SUP packages Requirements : You have a machine with Fedora 19 64bits installed (get fedora 19 64bits). A network access on Internet is necessary. A user with administrative rights (with sudo or root) is necessary. You need wget and unzip programs (yum install wget …

View page »

Simple test of CAR with recorded video

Testing on an existing video In order to test the CAR application, an RGB-D video is available at the following link :2013-09-10.zip If you do not have a configuration folder you can for example use the one provided for the simple CAR testing with live video in the previous webpage: SimpleConfigFile The video is mainly …

View page »

Install CAR Component on Fedora 17

Summary : Install ASUS Sensor and OpenNI Install PostgreSQL Database Install CAR package for the first time Test and use the CAR component Requirements : You have a machine with Fedora 17 64bits installed. (get fedora 17 64bits) A network access on Internet is necessary. A user with administrative rights (with sudo or root) is …

View page »

Check your CAR component installation

Summary : Check the Asus sensor and OpenNI installation. Check the Database installation. Check the SUP installation. 1 – Check the ASUS Sensor and OpenNI installation …by running OpenNI Viewer. A – OpenNI installed with packages : You must have openni-examples installed, run the following command to check : rpm -q openni-examples Then, you should …

View page »

Update CAR Component

Get the last version of SUP for Fedora 19 (x86_64) or SUP for Fedora 17 (x86_64) (also available by SUP RPM archives). If you have already installed SUP (Scene Understanding Platform) libraries of the CAR component and you want to udpdate CAR with a more recent version, you need to remove this previous RPM. 1 …

View page »

Test and use the CAR component

Summary : Record a video Display a video Simple test with pre-recording data Kreat-tool application Dem@care application 1 – Record a video Create the work folder : cd ~ mkdir asusdata Run the following command to record a data with the ASUS Sensor : cd /usr/local/inria/sup/bin ./sup-3DSensor_writer ~/asusdata/ If you want sava the data in …

View page »