Traffic Sign Board Recognition and Voice Alert System using Convolutional Neural Network

• Python
• Machine Learning

Since automobiles have become an indispensable medium of transportation, assurance of safety has been implemented in every country through proper road rules and regulations. Among them, traffic signs provide valuable information to the drivers and help to communicate the rules to be followed in that specific area. The purpose of a traffic sign is to convey a message quickly and accurately with minimum reading skills. Negligence, lack of attention, lack of familiarity, accidentally or deliberately not noticing traffic signs, distracting driving behaviors have been discovered as major reasons for the ignorance of road signs among the drivers which eventually lead to road accidents. Furthermore, drivers in un-urbanized communities may find it difficult in decoding the message conveyed by a specific road sign due to a lack of familiarity with the plenty of road signs in urbanized areas. Some drivers tend to ignore certain traffic signs believing that they are not necessary. Different attitudes of the drivers are also a reason for this ignorance. Ignorance or unfamiliarity with traffic signs could result in severe accidents and may even cost lives. To address the above problems, this project provides a method to detect and recognize traffic signs in real-time with higher accuracy and narrating the signs to the drivers. A system of this type can be used in both vehicle assistive systems and autonomous vehicles. The system is implemented using the Convolutional Neural Network (CNN) model.