Published Paper Details:

A very important aspect of the development of smart transportation networks and road safety is 'traffic sign recognition'. This survey covers all types of learning techniques applied to traffic sign recognition with respect to effectiveness, challenges, and progress. The study begins with an overview of the foundational methods of traffic sign recognition: traditional image processing approaches and their limitations. Further below, the heart of the survey is modern learning techniques, especially 'machine learning (ML)' and 'deep learning (DL)' approaches. The report reviews employ learning techniques applied in 'traffic sign recognition systems'. It discusses both the most recent developments in deep learning and advanced machine learning as well as conventional image processing. The review ranges from classical techniques such as colour and shape-based recognition to modern techniques such as 'Convolutional Neural Network (CNN),' which have significantly increased recognition accuracy. Several important datasets on which the training and evaluation should be done have been reported. One of them is 'German Traffic Sign Recognition Benchmark', and the challenges in practical applications along with the future research directions in this field have been highlighted. An overview has been provided regarding the effectiveness and evolution of learning techniques to improve 'Traffic sign recognition systems'.