logo

Select Sidearea

Populate the sidearea with useful widgets. It’s simple to add images, categories, latest post, social media icon links, tag clouds, and more.
hello@youremail.com
+1234567890
 

Siba Project

Highly sophisticated platform for urban parking services

Road recognition from satellite imagery

Location

SP/UK

Industry

Telecom

Team

+12 experts

Team Location

Spain (Madrid)

Challenge

Create large scale and highly sophisticated platform composed by several systems for providing urban parking services by using satellite data information over low resolution image recognition model.

Road recognition from satellite imagery has become a hot research field in recent years. Commonly is used in the city planning, cartography and to update previously detected roads in Geographic Information Systems (GIS) environment. From a simplistic perspective, at Siba we are mainly interested in extracting road-bounds inside cities and classified them by using neural networks.

Client goals

  • Define the mathematical algorithm for modelling KA System by using supervised and non-supervised machine learning models
  • Create general system .NET architecture for all integrated sub-systems:
    • Algorithm Foundation,
    • iPark Framework,
    • iPark Services,
    • Siba Administration,
    • and Third-Parties Portal.
  • Create a service-oriented architecture for allowing all third parties and devices to consume kernel logic
  • Create a powerful and efficient Backoffice web application tool for coordinating all sub-systems
  • Consolidate dynamic and automatic reporting system

Solution delivered by Iridium

Iridium mathematicians started by creating SVM (Support Vector Machine) classifications before applying specialized and patented image enhancement for establishing first plausible approaches over Quickbird-like images, which provide 2.44 meter detail in multispectral mode (blue, green, red and near-infrared bands) and 0.61 meter in panchromatic one.

From here two development teams were engaged for working in parallel in consolidating kernel architecture. After the successful completion of the first PoC´s and Kernel architecture, our customer decided to boost development team to over 12 engineers who work according to Scrum in 5 different teams.

Some technologies:

Tessellations, R, Python, TensorFlow, MathLab, Framework 4.5, N-Layered DDD Architecture, C#, ASP.NET MVC, WebAPI and SOA under WCF, Entity Framework, AngularJS, jQuery, , SQLServer, XML, XSLTs, XSD, Design Patterns, IoC (Dependency Injection), Microsoft Unity 2.1, TDD, Galio for Testing, .NET Instrumentation for Monitoring, WCF Custom Message Security, etc..