Business Areas
GeoSystem Research, a global leader in Earth science and technology
Business Areas
Artificial Intelligence-based Technologies
Business Introduction
Artificial intelligence (AI) technology is a technology that uses data-based models that are created from vast amounts of accumulated data. GeoSystem Research has researched and developed various key technologies in the marine and fisheries sectors using AI technology, such as real-time marine forecasting (temperature, sea level, Chl-a), red tide forecasting, satellite image super resolution, and missing spatial data imputation. In addition, we are working on the development of various AI-based marine and fisheries-related key technologies using the marine big data platform we have built.
Our Technology
- Time series prediction model using DNN, RNN(LSTM)
- ConvLSTM, Predrnn++ based 2D spatiotemporal prediction model
- GAN-Resolution-based spatial data resolution upscale model
- GAN-Inpainting missing data imputation model
Applications
- Real-time ocean forecast data production (water temperature, cold water, Chl-a, etc.)
- Real-time marine information related to fishermen and fishery activities
- Missing data imputation and super resolution of satellite data
- Production of red tide prediction data
.png)
yeongmin Park, Director
Tel. 031-5180-5871
E-mail. ympark@geosr.com
.png)
Business Areas
-
.jpg)
Satellite image processing and analysis
To monitor and evaluate natural or artificial phenomena, data can be collected directly from the site (in situ) or remotely from satellites. Remote sensing is highly utilized because it can measure wide-area information almost simultaneously and is relatively inexpensive. GeoSystem Research produces baseline data on oceanic phenomena using remote sensing technology and provides technical support and solutions in applicable fields. We also provide solutions for an integrated management, from selecting appropriate solutions in the event of marine environmental problems, collecting data, processing field data and analyzing digital images, developing algorithms, to creating thematic maps -
Automated Seismic Data Processing
Seismic survey is a survey method that explores the geological features of deposits or bedrock by structure of a rock formation or bedrock by sending sound waves and measuring the transmission time and strength of these sound waves at a number of points. Since processing of seismic data depends on the capabilities of technicians, various methodologies, sound source characteristics, and hydrophone array, and work-flow, a considerable amount of time and money is spent on data processing to obtain high-quality data. At GeoSystem Research, we have developed and are operating an automated data processing program by building a data processing module and an optimal work-flow that takes into account the sound source characteristics and hydrophone array for shallow water environments such as coasts, rivers, and estuaries. Using this program, we can process large amounts of data several to several dozen times faster, saving time and money, and outputs high-quality data that is easy to interpret. Currently, the program has a total of 20 data processing modules and is being continuously updated. -
CCTV Image Processing and Analysis CCTV
GeoSystem Research conducts real-time video monitoring of major erosion-affected coasts in Korea to obtain long-term continuous data on coastline changes. By identifying causes of erosion using our CCTV image processing and analysis technology, we provide baseline data needed to establish primary coastal maintenance plans and coastal maintenance projects. -
Computer Vision-based Marine Organism Detection
The Korean government has been promoting the application of big data, IoT, AI in the ocean, fisheries, and shipping sectors to meet the increasing demand for aquaculture and follow the Fourth Industrial Revolution trend. GeoSystem Research provides various computer vision-based technical solutions, such as automatic marine organism (ex. fishes and harmful marine organisms such as jellyfish and salpa) detection and warning system, smart fish farm monitoring, and aquatic ecosystem survey support. We are continuously researching to provide the latest, optimal computer vision-based solutions that enables our clients to respond to situations that needs urgent measures. -
 맞춤형 수산정보 서비스 시스템 개념도 2(0).png)
Big data
Big data is a technology that extracts value from data, including large structured or unstructured data sets of tens of terabytes in size, that exceeds the capabilities of existing database management tools, and analyzes the results. It is important to not only collect and organize data, but also to find meaning among the data and produce valuable information from it. GeoSystem Research is leading the field of convergence technology that combines artificial intelligence and advanced ICT technologies with marine forecasting. We are aiming to lead the future of marine technology by establishing a framework for providing marine forecasting information, including distributed data warehouse, RDBMS data linkage, big data integrated management system, and construction of unstructured data DB and ontology. -
Digital Offshore Windfarm Site Information Map
For clients planning to construct offshore wind farms, we provide scientific data for environmental suitability, stakeholder acceptance, optimal site information map, site selection guidelines to support decision making. This also helps resolve conflicts between developers and local residents that may arise due to a lack of objective and scientific information. We create offshore windfarm site information maps using our marine environmental information (flow, wave, temperature, conductivity, wind, etc.) and marine spatial information. -
Unmanned Surface Vessel Observation
Until recently, fixed observation platforms (e.g., observation buoy, fixed observation station) or manned vessels have been used to monitor water quality and physical characteristics of oceans, rivers, and lakes. With recent advances in autonomous navigation technology, position control technology, and data transmission technology, it is now possible to conduct three-dimensional observations while moving to various points using unmanned surface vessel (USV) platforms in remote areas. Our USV platform can be used to quickly respond to changes in the water environment, such as localized pollution incidents which can’t be detected from a fixed location, and to conduct observations at various points in a short period of time. The USV can collect data at multiple depths for each survey points using a winch, and the collected data is used to derive images showing three-dimensional distribution of the water environment, which helps surveyors to analyze the water environment intuitively. In addition, a side-scan sonar, multi-beam sonar, and LiDAR device can be attached to the hull to conduct underwater and on-water topographic surveys while maintaining a constant speed along pre-planned routes in ocean and river environments. -
Unmanned Aerial Photogrammetry and LiDAR Survey
The traditional method of conducting a beach profile and topographic survey has the disadvantage of being inefficient, as a large number of people are required to work for a long period of time, and there is a risk of safety accidents when surveying cliffs, rocks which does not cover, or rocks which covers and uncovers. GeoSystem Research is currently carrying out a national geospatial information project that combines unmanned aerial vehicles with the latest sensors, a method which ensures the safety of surveyors while collecting high-precision geospatial information data, increasing work efficiency. The high-precision spatial information built this way is used to establish risk assessment technology for complex coastal disasters by building models and planning disaster prevention measures. Our geospatial information is also widely used for monitoring changes and current conditions of coastal areas and coastline database building projects. -
Real-time Ocean Forecast System
The real-time ocean forecast system uses observation data and numerical models to predict ocean phenomena, such as water level, flow, water temperature, conductivity, waves, surge height, diffusion, etc., in real time. GeoSystem Research has built in-house a real-time ocean forecast system that simulates ocean circulation and waves to produce forecast data for the Pacific Northwest, Northeast Asia, the Korean Peninsula, and other major coastal areas. This forecast data can be used in various ways, such marine environmental forecast regarding climate change, SAR (search and rescue) and disaster prevention activities in the event of a marine accident, and military operations. We have also built a real-time thermal effluent dispersion prediction system, a red tide movement and diffusion prediction model, and a visualization system, and we are committed to research and development of the latest technologies, such as improving the accuracy of forecast data using the latest data assimilation techniques.