Comparison of Vehicle Categorisation by Convolutional Neural Networks using MATLAB

The number of cars driving on the roads is increasing every year. Due to the large number and especially the weight of passing vehicles, roads are in critical condition and need more and more investments for their repair. This incurs significant costs for road operators. Article goals are the visual classification of passing vehicles over the tested platform using neural networks. The differences, performance and their structures comparisons of these two neural networks, AlexNet and GoogLeNet, will be described.