iOS, Objective-C, C, C++, Android SDK, Java, Tesseract OCR, OpenCV, ASP.NET, MVC, Entity Framework, MS SQL Server, Bootstrap, jQuery, WordPress, Google Maps API
Mobile development / Business applications
With a team of individuals with more than 25 years’ experience serving the diplomatic corps, the client was aware that consulates face increasing pressure to screen travellers before entry and are constrained by dwindling resources.
The scope of the application that the client had envisioned was to help both applicants and consulates, by enabling a remote and much faster application process and making applicant engagement more secure.
The idea was to combine multi-channel, applicant engagement features and “score” an application before “judgement” by the consular corps.
The client had heard of Romania’s reputation as a convenient outsourcing destination, with several providers that develop high quality software for affordable rates, and decided to narrow down his search to our country. After screening multiple software development companies, he chose to work with us thanks to our mobile development experience and machine learning capabilities.
The most challenging aspect that we managed to overcome during implementation was performing OCR in various light conditions, from under-exposed to over-exposed passport details.
Problems stemmed from the fact that some of the mobile devices used for photographing the users’ passports were not equipped with flash options and the photos were low – resolution.
The image processing algorithms that we wrote in C++ successfully eliminated the lack of clarity in the images.
The main technologies involved in implementing the mobile application for scanning passports were:
iOS with C and Objective-C - for the iPad application
Android SDK with C and Java – for the Android application
C++ - to write an image processing algorithm for the application
Tesseract OCR - one of the most popular open source optical character recognition (OCR) libraries, "to read" the information in the photos
The machine learning algorithm interprets the information in the photos and fills out the application form with the data corresponding to each field
OpenCV - one of the most popular libraries for computer vision tasks, was used for both mobile apps
The gov.uk API – to validate the companies that the applicants claimed to work for
For developing the Applicant & Manager websites: ASP.NET with MVC, Entity Framework, MS SQL Server, Bootstrap and jQuery. The presentation website was built on WordPress. Additionally, integration with Google Maps facilitate the validation of some of the information that the users entered in their visa application, such as their current and future address and the address of their employer
Access to machine learning expertise, which is generally very hard to find, especially in Romania
Access to a dedicated development team of senior mobile developers, with 5+ years’ experience in iOS and Android
Timely deliveries: the first release was ready a month earlier than the client expected
Excellent communication skills and flexibility
There is no better partner than Roweb. Their large arsenal of competencies exceeded our requirements. The implementation team is confident, consistent and reliable, delivering perfectly to specification every time and far ahead of schedule. The developers are incredibly positive, adept, creative and artful, pouring heart and soul into a project.
Paul Horner
Director at DigiConsul GmbH