Performance Analysis of Vehicle Classification System Using Type-1 Fuzzy, Adaptive Neuro-Fuzzy and Type-2 Fuzzy Inference System

An algorithm is developed to find the accuracy of the system for vehicle classification using different techniques and it is expected that it may further improve the results as imperfection and uncertainty in the vehicle data are very nicely handled by type-2 fuzzy system.

Abstract

Vehicle Class is an important parameter in road traffic measurement. In this paper authors developed an algorithm to find the accuracy of the system for vehicle classification using different techniques. The algorithm mainly reads the inference system and applies various input samples, check the class of each sample and calculate the accuracy. Initially the classification was done using Type-1 fuzzy logic system and found that the accuracy of the system was not acceptable. To increase the accuracy there was a need to meticulously adjust the shape and placement of membership function of different input variables. This process was time consuming and inaccurate. Then the same objective was implemented using adaptive neuro-fuzzy inference system and it was observed that the membership functions are finely tuned by anfis and accuracy was greatly increased. Finally, type-2 fuzzy inference system is used for the same purpose and it is expected that it may further improve the results as imperfection and uncertainty in the vehicle data are very nicely handled by type-2 fuzzy system.