To improve monitoring sand transport, and in an effort to curb illegal mining and transport, the International Institute of Information Technology, Hyderabad (IIITH) has developed a solution equipped to tackle the problem of identifying lorries with non-standard number plates.

Vahan Eye, according to IIITH, was developed after the Information Technology department approached them seeking a system for the Telangana Mineral Development Corporation (TGMDC) that would automatically recognise registration number plates. The system required would have to be reliable, but cost effective as well.

While standard registration plates are easy to identify, given that many lorries have hand painted registration number plates, or plates that vary widely in terms of font and font size, the challenge was for the system to recognise such number plates. Commercially available registration plate recognition systems in the market, according to IIIT-H, were expensive.

With these requirements, IIIT-H Adjunct Faculty member Verra Ganesh Yalla, who is also adjunct faculty, and his team used prior research undertaken by IIIT-H’s Centre for Visual Information Technology and worked on this. His team then developed methods to implement it by working on an already existing registration plate recognition system. The team analysed the workflow and improved upon the handwritten character recognition system to make it more reliable.

What sets their work apart is that they integrated the analytics as a plug-in into an opensource platform.

As part of their pilot project, Vahan Eye was used along the Vijayawada-Hyderabad highway in Chityal. The IIIT-H team put systems in place here, including installing cameras. By means of Vahan Eye system, they then tracked lorries entering Telangana and cross checked them against a list containing details of 40,000 approved lorries. The team then made modifications to suit the requirements of the TGMDC.

Since September last year, the system has been operating without interruption. Even with difficulties such as poor lighting at night and festival decorations like garlands covering number plates, the algorithm has performed reliably and is continuing to improve as it processes live data.

In September, the TGMDC had announced their intention of calling for tenders for the solution. The move was in consonance with the State’s sand mining policy that seeks to bring about transparent sales and properly regulated transportation of sand. While vehicles carrying sand from quarries to stockyards are tracked using GPS, the new ANPR system is designed in such a fashion that it brings about checks by capturing the number plates of passing vehicles, including those transporting sand.


Leave a Reply

Your email address will not be published. Required fields are marked *