A Novel Convolutional Neural Network Classifier for Real-Time Traffic Sign Detection

An automated system for the detection and classification of traffic signs using Convolutional Neural Networks (CNNs) is presented in this paper. Enhancing traffic management, road safety, and the advancement of self-driving cars are the goals of this research by accurately identifying and classifying traffic signs. The CNN-based approach offers a practical choice for trustworthy and efficient traffic sign recognition. By putting this into practice, we can advance road safety, traffic control, and autonomous vehicle technology. In order to progressively extract complex features, stacking layers are utilized. The proposed CNN based system's real-time detection and classification of traffic signs enhances the development of autonomous vehicles, road safety and traffic control. By displaying efficient traffic sign's image detection and classification, this research showcases accuracy and simplicity in CNN model training. Reduction in weight and dropout are regularization strategies that stop overfitting for better results on training and testing data. Convolution, Activation, Maxpooling 2D, and Flatten layers are crucial CNN layers. It accepts input formats such as images, videos, and supports real- time detection. In order to obtain a minimal loss and acceptable training accuracy, the CNN model needs to run through more epochs.