iTag

Semantic enhancement of Earth Observation data

iTag is a library to tag a footprint with the following information :

• continents
• countries
• cities
• regions and states
• geophysical plates
• volcanoes
• land cover (i.e. forest, water, urban, cultivated, herbaceous, desert, snow, flooded)
• population count

You can access an online instance [here] (http://mapshup.com/projects/itag) as a web service.

See [video capture of itag applied to Pleiades HR and Spot5 images database] (http://vimeo.com/51045597) and access trough [mapshup] (http://mapshup.com/projects/mapshup)

iTag is used by [RESTo - REstful Semantic search Tool for geOspatial] (http://github.com/jjrom/resto)

Installation

We suppose that $ITAG_HOME is the directory containing this file.

Prerequesites

• PHP (v5.3+) command line
• PostgreSQL (v9.0+) with unaccent extension
• PostGIS (v1.5.1+)
• GDAL (v1.8+) with python support (for land cover preparation only)

Note: iTag could work with lower version of the specified requirements. However there is no guaranty of success and unwanted result may occured !

Configure postgresql

Edit the PostreSQL postgresql.conf and be sure that postgres accept tcp_ip connection.

    # Uncomment these two lines within postgesql.conf
    listen_addresses = 'localhost'
    port = 5432

Step by step

1. Get data

    # Note : $ITAG_HOME **must be** an absolute path (not relative !)
    git clone https://github.com/jjrom/itag-data.git $ITAG_DATA
    cd $ITAG_HOME/_install
    unzip $ITAG_DATA/data.zip
   

2. Install database

    # Note : "password" must be the same as 
    # the value of 'password' parameter in $ITAG_HOME/include/config.php
    
    $ITAG_HOME/_install/installDB.sh -F -d <path_to_postgis_directory> -p password
   

3. Populate database

    # 
    # Note : Read this if you are using Fedora, Red Hat Enterprise Linux, CentOS,
    # Scientific Linux, or one of the other distros that enable SELinux by default.
    #
    # SELinux policies for PostgreSQL do not permit the server to read files outside
    # the PostgreSQL data directory, or the file was created by a service covered by
    # a targeted policy so it has a label that PostgreSQL isn't allowed to read from.
    #
    # To make the following scripts work run the following command as root
    #
    #   setenforce 0
    # 
    # Then after a successful execution relaunch the command
    #
    #   setenforce 1
    #
    # General datasources
    $ITAG_HOME/_install/installDataSources.sh -F -D $ITAG_HOME/_install/data
    
    # Geonames
    # First you need to download geonames data in $GEONAMES_DIR directory
    export GEONAMES_DIR=/a/temporary/directory
    cd $GEONAMES_DIR
    wget http://download.geonames.org/export/dump/allCountries.zip
    wget http://download.geonames.org/export/dump/alternateNames.zip
    wget http://download.geonames.org/export/dump/countryInfo.txt
    wget http://download.geonames.org/export/dump/iso-languagecodes.txt
    unzip allCountries.zip
    unzip alternateNames.zip
    # Remove unwanted comment from countryInfo.txt
    grep -v "^#" countryInfo.txt > tmp.txt
    mv tmp.txt countryInfo.txt
    $ITAG_HOME/_install/installGazetter.sh -F -D $GEONAMES_DIR
   
    # Wikipedia
    # This step is optional and can only be performed if you have the geolocated wikipedia data (which probably you don't have :)
    # In case of, these are the steps to follow in order to install this database within iTag
    #
    # Put the geolocated wikipedia data in $GEONAMES_DIR/wikipedia directory, then run the command
    #
    # $ITAG_HOME/_install/installWikipediaDB.sh -D $GEONAMES_DIR/wikipedia
   

4. Precompute landcover

    Go to ["Global Land Cover 2000" global product] (http://bioval.jrc.ec.europa.eu/products/glc2000/products.php) and download glc2000 GeoTIFF file
   
    $ITAG_HOME/_install/computeLandCover.php -I path_to_glc2000_tif_image
   

5. Deploy application

    $ITAG_HOME/_install/deploy.sh -s $ITAG_HOME -t $ITAG_TARGET
   

Note : depending on your server performance, the landcover computation can take a long time (more than two hours)

Using iTag

We suppose that $ITAG_TARGET is accessible to http://localhost/itag/ in Apache.

To tag footprint on Toulouse with geophysical information and all cities with a pretty GeoJSON output, open this url within you browser

    http://localhost/itag/?geophysical=true&countries=true&cities=all&pretty=true&footprint=POLYGON((1.350360%2043.532822,1.350360%2043.668522,1.515350%2043.668522,1.515350%2043.532822,1.350360%2043.532822))

Available parameters for Web service are :

• &continents=true
• &countries=true
• &cities=main (or &cities=all)
• &geophysical=true
• &regions=true
• &landcover=true
• &french=true
• &hierarchical=true
• &ordered=true
• &pretty=true

You can check this [running instance] (http://mapshup.com/projects/itag/)

Examples :

Tag footprint on Toulouse with geophysical information and all cities with a pretty GeoJSON output

    http://mapshup.com/projects/itag/?geophysical=true&countries=true&cities=all&pretty=true&footprint=POLYGON((1.350360%2043.532822,1.350360%2043.668522,1.515350%2043.668522,1.515350%2043.532822,1.350360%2043.532822))


Tag footprint intersecting France, Italy and Switzerland with cities, regions and states. Hierarchical result as pretty GeoJSON output

    http://mapshup.com/projects/itag/?regions=true&hierarchical=true&ordered=true&countries=true&cities=all&pretty=true&footprint=POLYGON((6.487426757812523%2045.76081241294796,6.487426757812523%2046.06798615804025,7.80578613281244%2046.06798615804025,7.80578613281244%2045.76081241294796,6.487426757812523%2045.76081241294796))