In order to meet the development of The Times and enhance the intelligence of new energy vehicles, and solve the problem that the traditional image processing is too demanding for the feature quantity and the recognition effect is poor, a road recognition research based on deep learning is designed. Four deep learning networks (VGGNet, ResNet, GoogLeNet and AlexNet) were used in the study. A large number of images of different road surfaces were collected as experimental data, and the data were trained by the deep learning network. By comparing the results of four kinds of networks, the network which can be used in intelligent vehicles and has better recognition rate is finally analyzed. The research achieved the expected effect and laid the foundation for the subsequent design.

Fei Xie, Li Yang, Hailong Duan. Pavement Recognition Based on Multiple Deep Learning. Academic Journal of Computing & Information Science (2022), Vol. 5, Issue 11: 90-96. https://doi.org/10.25236/AJCIS.2022.051114.

[1] Xiao Wangxin, Zhang Xue, Huang Wei, et al. A New Algorithm for Automatic Identification of Pavement Damage [J]. Highway Traffic Technology, 2005, 3(11): 79-82.

[2] Shi Weijia, Ma Shui, Ma Bin, et al. Design and implementation of pavement recognition system [J]. Automation instrumentation, 2011, 32(04): 62-65.

[3] Zhang Xiaolong, Sun Renyun, Liu Changwei, et al. Research on road surface recognition method based on road surface characteristic coefficient [J]. China test and measurement, 2015, 41(08): 31-35. 

[4] Gao Wei. Research on Road Recognition Technology of Mobile Robot [D]. Hebei: Hebei University of Technology, 2014.

[5] Jing Weipeng, Zhang Mingwei, Lin Jingbo. A neural network framework search method for remote sensing image classification [J]. Journal of Harbin university of science and technology, 2021, 26(01): 25-29. 

[6] Lu Haoyuan, Yu Lu, Zhou Xingyu, et al. A survey of semi-supervised deep learning image classification methods [J]. Computer science and exploration, 2021, 15(03): 1038-1048.

[7] Zhang Xiaoling. Study on epilepsy prediction based on deep learning and scalp electroencephalogram [D]. Tianjin: Tianjin Vocational and Technical Normal University, 2022. 

[8] Zhang Ke, Feng Xiaohan, Guo Yurong, et al. A survey of deep convolutional neural network models for image classification [J]. Journal of image and graphics, 2021, 26(10): 2305-2325.

[9] Li H, Chen D. Complexity and Accuracy Analysis of Convolutional Neural Networks in Scene Text Detection [C]. International Conference on Electronic Information Technology and Computer Engineering (EITCE), 2019: 416-1419. 

[10] Li Y D, Hao Z B, Lei H, et al. A survey of Convolutional Neural Networks [J]. Journal of Computer Applications, 2016, 36(9): 2508-2515.

[11] Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks [J]. Advances in neural information processing systems, 2012, 25: 1097-1105. 

[12] Wang Z, Zheng X, Li D, et al. A VGGNet-like approach for classifying and segmenting coal dust particles with overlapping regions [J]. Computers in Industry, 2021, 132: 103506.

[13] Rajan S, Poongodi C, Devaraj S, et al. Novel Deep Learning Model for Facial Expression Recognition based on Maximum Boosted CNN and LSTM [J]. IET Image Processing, 2020, 14(7): 1373-1381. 

[14] Xie S, Zheng X, Chen Y, et al. Artifact Removal using Improved GoogLeNet for Sparse-view CT Reconstruction [J]. Scientific Reports, 2018, 8(1): 1-9.

[15] Wen L, Li X, Gao L. A transfer convolutional neural network for fault diagnosis based on ResNet-50 [J]. Neural Computing and Applications, 2020, 32(10): 6111-6124.

[16] Zhao Jin. Theory and Application of Sparse Deep Learning [D]. Xi 'an: Xidian University, 2019.

[17] Ren Jinjun, wang Ning. Research on loss function in artificial neural networks [J]. Journal of gansu normal university, 2018, 23(02): 61-63.