Welcome to Francis Academic Press

International Journal of New Developments in Engineering and Society, 2023, 7(3); doi: 10.25236/IJNDES.2023.070305.

Research on Data Enhancement Method for Complex Pavement


Tingting Zhang, Shuyan Ren, Hailong Duan, Shuoyi Wen

Corresponding Author:
Shuyan Ren

School of Automation and Electrical Engineering, Tianjin University of Technology and Education, Tianjin, China


In the era of intelligent vehicle development, environmental awareness technology as the basis of vehicle decision-making and route planning, affected by the weather, accurate road image recognition is particularly important. Aiming at this problem, this paper constructs a large scale pavement image dataset and proposes an improved image generation method based on CycleGAN data enhancement. The core innovation of the method is the improvement of the model structure, introduce Wassertein distance, and implement global Lipschitz constraint with spectral constraint. The experimental results demonstrate that the accuracy of the data set is up to 91.85%, which is 6.41% higher than the original CycleGAN, and the convergence speed is faster.


Pavement Image, Dataset, CycleGAN, Image Generation

Cite This Paper

Tingting Zhang, Shuyan Ren, Hailong Duan, Shuoyi Wen. Research on Data Enhancement Method for Complex Pavement. International Journal of New Developments in Engineering and Society (2023) Vol.7, Issue 3: 28-35. https://doi.org/10.25236/IJNDES.2023.070305.


[1] Russo Fabrizio. An image enhancement technique is combining sharpening and noise reduction [J]. IEEE Transactions on Instrumentation and Measurement, 2001, 51(4):824-828. 

[2] Mao Xudong, Li Qing, Xie Haoran, et al. Least squares generative adversarial networks[C]. Proceedings of the IEEE international conference on computer vision, 2017:2794-2802. 

[3] Arjovsky Martin, Chintala Soumith.Bottom Léon. Wasserstein generative adversarial networks[C]. International conference on machine learning, 2017:214-223. 

[4] Brock Andrew, Donahue Jeff, Simonyan Karen. Large scale GAN training for high fidelity natural image synthesis[C]. International Conference on Learning representations, 2019. 

[5] Collins B, Deng J, Li K, et al. Towards scalable dataset construction: an active learning approach [C]/European conference on computer vision. Springer, Berlin, Heidelberg, 2008: 86-98.

[6] Lin Jenn L., Dai Xingyuan, Li L., et al. A new frontier in artificial intelligence research: generative adversarial networks[J]. Journal of Automation, 2018, 44 (5): 775 -792

[7] Wang Kunfeng, Gou Chao, Duan Yanjie, et al. Research progress and prospect of generative adversarial network GAN [J]. Journal of Automation, 2017, 43 (3): 321 a 332

[8] Wang Gongming, Qiao Junfei, Wang Lei a generative adversarial network in the sense of energy function [J]. Journal of Automation, 2018, 44 (5): 793 I 803

[9] Yan Bei and Zhang Jianlin. Study on image data generation based on spectral constraint for generative adversarial networks [J]. Semiconductor Photonics, 201940 (6): 896 -901

[10] Tian L, Zheng Y, Cui Q. Research on Data Enhanced Ancient Pictogram Recognition Method Based on Convolutional Neural Network[C]// the 2019 3rd High Performance Computing and Cluster Technologies Conference. 2019.

[11] Peng Peng. Image style conversion based on CycleGAN [D]. Chengdu:University of Electronic Science and Technology, 2019.

[12] Guo Jingda, Ma Xu, Sansom Andrew, et al. Spanet: Spatial pyramid attention network for enhanced image recognition[C]. IEEE International Conference on Multimedia and Expo, 2020. 

[13] Cai Zhiling, Weng Qian, Ye Shaozhen, et al. Scene classification of high-resolution remote sensing images based on Inception V3 model [J]. Remote Sensing of Land Resources, 2020, 32 (3): 80-89