Automated vehicle classification in India

Date: 4/3/2010

Source: ‘Passage through India’, Tolltrans 2010, pg. 52

A new Automatic Vehicle Classification (AVC) system, developed by Tecsidel with software development from Ashoka Technologies, has been deployed in India as a part of toll road project managed by Ashoka Buildcon.

In India, the National Highways Authority of India (NHAI) divides vehicles into four toll classes. There is a diverse variety of vehicles using India’s highways, so any AVC system needs to shrewd at classifying a vehicle that it may not have many examples of in its database.

The improved AVC system, designed specifically to perform well with the Indian classification system, is based on double-wheel detection and 3D vehicle profiling, leading to an improved and more accurate vehicle classification. Vehicle profiling focuses on classification of two-axle vehicles by means of an overhead laser scanner and intelligent shape-recognition software. Double wheels are identified with use of oblique treadles (mechanical switches that are actuated by the vehicle’s tyres). Thanks to this latest technology the number of vehicle classification errors was reduced by 80%.

Keywords: Identification, Payment