Skip to content

GEOMATRIX UAB: Extreme Geo-computing

GEOMATRIX UAB is a small private Lithuanian company, focusing on applied research, development of technological solutions, consultancy in geo-informatics and providing services of automated processing of large amounts of spatial data. It was originally designed to become a GMES Service Provider, focusing on innovative satellite data processing solutions.

The main business idea of the company is development and implementation of “robotic” automated processing work-flows, capable of processing large amounts of data in short time and with low human interaction – therefore with high efficiency and at a low cost – taking advantage of remotely-operated computing systems, or even “external boxes” established for the clients and integrated into their technological work-flows. This is why the company positions itself as the one capable of doing “extreme” geo-computing.

GEOMATRIX UAB is an advanced user and passionate advocate of FOSS (Free and Open Source Software), trying to develop efficient and competitive industrial geo-computing solutions without using any commercial software. We have established an ergonomic, but highly-productive true 64-bit multi-processing environment, based on Linux OS, GRASS GIS, GDAL/OGR libraries, etc. System re-programming techniques are used to combine Linux shell components and utilities with professional GIS/RS software functions, libraries and tools, binding them into automated data processing work-flows, designed to utilize full computing power down to the last processor core or bit of memory.

The first successful project of the company was development of an operational mapping solution for a Swiss company PASSAM AG for processing and representation of air quality monitoring data collected in Lithuania by using passive sorbents sampling method. Thousands of operational maps were produced by automated scripts and delivered to the clients as digital images via FTP or on-line repositories. Automated operational mapping approach can be used in a broad range of projects for representation, analysis and quality control of large thematic datasets collected in many sampling sites – an ideal solution for large environmental monitoring networks, emergency response services, field data collection and analysis, etc.

Development of automated parallel computing work-flows for processing of large amounts of vector and raster datasets, particularly high and very high resolution satellite imagery, is currently the most important activity of the company. Back in 2011, we developed a series of automated work-flows for extraction of water bodies from IMAGE 2006, which was successfully applied by the Institute of Aerial Geodesy (Lithuania) for the production of GMES Reference Datasets of hydrographic network over 38 European countries. A year later we started the production of high resolution layers of Water and Wetlands from high resolution satellite imagery over the EEA39 territory for the EEA within a GMES Initial Operations (GIO-Land) project . Both projects are coordinated by Spanish company Indra Espacio SA.

The company has developed a series of efficient production work-flows for automated pre-processing of standard satellite imagery products (data format/type conversion, re-projection, ortho-rectification, mosaicking, extraction/manipulation of bands, pan-sharpening, metadata extraction, thumbnailing, renaming, packaging, cloud masking/patching). Although rather simple, such routine operations may require a substantial amount of time and resources, if carried out manually on a larger number of images. Therefore automated pre-processing of satellite imagery has become a standard practice. Experience in this field allowed the company to take part in an FP7 project LAMPRE, currently developing a GMES downstream services on landslide modeling, vulnerability assessment, preparedness and recovery management.

Examples of products

Illustration of the RapidEye image pan-sharpening work-flow: left image – the original RapidEye level 3A product (fragment of Panevėžys town in Lithuania) taken on 2011-08-27 by RE-1 sensor, shown in RGB-542 false color composit; middle image – panchromatic image with 1 m pixel resolution produced of a series of color aerial photo images with 0.5 m pixels; right image – pan-sharpened RapidEye image produced in 1 m pixel resolution and RGB-541 false color composit

A series of accumulative biophysical products of Slovakia made in a full-scale IRS-6 resolution. Source: IMAGE2006/2009/2012

The main projects:

Data processing services:

Contact information:
Address: Technopolis, Europos pr. 121, LT-46339 Kaunas, Lithuania
tel.: +370-699-99940
fax.: +370-37-211382|