Adaptive Learning of Radial Basis Function Neural Networks Based on Traffic Sign Recognition using Principal Component Analysis

Using the PCA and RBF neural networks developed in this study, it was possible to develop a practical method for recognising traffic signs. PCA has been used in traffic sign recognition algorithms for several years. It is among an autonomous driving system's most prevalent image representation techniques. The picture is not only reduced in dimensionality, but some of the fluctuations in the digital image and the image data are retained. It is true that when PCA was completed, the RBF neural netts' hidden node neurones were modelled using the training images' intra-class discrimination qualities in the hidden layer neurone. RBF neural networks benefit from this because it allows them to acquire a wide range of changes observed in the low-dimensional feature space, increasing their generalisation capabilities. The suggested approach is tested on different template traffic signs, with positive results. Results from the experiments demonstrate that the suggested technique has a promising recognition performance.