An Android based Quantum GIS prototype Ramon Carrillo, Daniel Ochoa dochoa@espol.edu.ec
Summary Background Past works Quantum GIS Mobile Quantum GIS Results Future works
Background Robotic and computer vision centre (CVR) CVR focused mainly on data acquisition, processing and interpretation. Research on image analysis (biological, human motion, 3D reconstruction). The only 100% FOSS centre at ESPOL. Supports KOKOA, ESPOL free software community.
Past works... no GIS Ocean currents Vehicle speed
Past works... no GIS Novel uses for remote sensing algorithms. Road detection Spatial relations
Current R&D interests CVR is the newest member of ESPOL's hydrology group: Need for Geo-referenced data collection. Extension and adaptation of HW and SW tools. Image analysis (pattern recognition). Limited image data in Ecuador.
A GIS tool prototype Field data collection using mobile computers (phones). Why a phone? Affordable. Decent CPUs + Sensors. Networking. Data calculation at input time. In-the-field geo-referenced data. Crowd-sourcing.
Our prototype Sensor network Data center Network Gateway Node Mobile GIS Phone OS HW User (gps,camera,etc)
Quantum GIS Recommended by polimi colleagues. Fully featured GIS package: Cross-platform (Linux, Windows, Mac) open GIS. Direct viewing of vector and raster data in different formats. Mapping and interactive exploration of spatial data. Create, edit and export spatial data. Supported by several institutions. Site: www.qgis.org
Quantum GIS DTM models Resource inventory Urban Planning
Quantum GIS architecture Core + extension mechanism (plug ins) QT (widgets) Plug-in Plug-in core GDAL/OGR Spatial DB WMS GIS Data Plug-in Sensor Data Core libraries include: QT (Graphic user interface) Native support for PostGis, Spatialite OGR (ESRI Shapefiles,S-57, SDTS, Oracle Spatial, and Mapinfo) GDAL (Raster, Aerial, Satellite data) Online spatial data (OGC-compliant), WebMapService (WMS).
Quantum GIS architecture Plug-ins written in C++ or python About 160 plug-in available: GIS Data source access (Grass,QspatialLite). Sensor data access (Garmin GPS). Layer manipulation. Statistics, etc.
OS selection Windows phone, rather new, proprietary software. IOS, proprietary software. Limited and expensive hardware choice. Android, runs on top of Linux. Not as free/open as it should be. Runs on many devices. SDK & NDK development.
Network node Mesh network using Zigbee protocol Xbee module (900 Mhz) attached to USB port. FTDI driver for serial data exchange over USB port. Included in Linux kernel. Qgis plug-in written in C. Data logging. Data plotting. video
Mobile QGis Based on Marco Bernasocchi's work Full Qgis desktop software migration. Version for android tablet. Main result: cross compiled version of core-libraries Better performance on a reduced software stack. Basic features should work with android VM: Data Access. Layer manipulation. GPS support. Funded by GSoC and ESPOL
Mobile QGis architecture Distribute tasks across the Android stack. Dalvik (JVM) SDK Java launcher + JNI QML NDK Qt GDAL GIS Data Spatialite Sensor Data
Mobile QGis challenges Rethink graphic user interface (new GUI). Consider phone hardware access restrictions. Get the correct set of development tools and utilities. Necessitas: Qt-android port. Ministro: Qt installer External libs (GDAL, GEOS, Proj4, expat, GSL, Iconv, QWT, sqlite, libpq, open-ssl) Cmake, Android ant build system.
Results QGis on Android consists of: The cross compiled dependencies (.so files) An library file (.so shared object) of QGIS (there is no executable binary). A QML-based graphic application. Software distribution: apk package, ministro takes care of QT libs. Gps support enabled. Radio communication implemented and tested.
Mobile QGis Rendering and Gps support
Mobile Qgis GIS data edition
Future works Additional Data Sources (WMS) Add Off-line support. Enable access to more phone's sensors. Compas Cammera,etc. NDK interface for Xbee module. Java interface for Bluetooth module.
Questions?