A Probabilistic Superpixel-Based Method for Road Crack Network Detection
This paper presents a probabilistic superpixel-based method for detecting road crack networks. The proposed method includes the techniques of skeletonization and end-growing at the superpixel level, which lend to the extraction of slender crack features from road images. Probabilistic crack pixel refinement is implemented, followed by geometry filters and binary crack cleaning operations, with the end goal of presenting cracks in their simplest form for further high-level characterization. The performance study used to characterize this crack detection algorithm was not constrained by crack type, pavement type, or even image resolution. This approach boasts a median pixel-wise distance error rate of less than one pixel, and for a 100-image dataset, the average detected crack length was within 18% of the ground truth crack length.
KeywordsImage processing Probabilistic methods Superpixels Geometry filtering Crack detection
The authors would like to acknowledge Murata Manufacturing Co., Ltd. for supporting the work presented in this paper.
- 1.Erickson, S.W.: Street pavement maintenance: road condition is deteriorating due to insufficient funding. Office of the City Auditor, San Jose (2015)Google Scholar
- 2.Vlacich, B.: State of the pavement 2016. Virginia Department of Transportation, Richmond, VA (2016)Google Scholar
- 3.Tanaka, N., Uematsu, K.: A crack detection method in road surface images using morphology. In: IAPR Workshop on Machine Vision Applications, pp. 154–157 (1998)Google Scholar
- 5.Chambon, S., Subirats, P., Dumoulin, J.: Introduction of a wavelet transform based on 2D matched filter in a Markov random field for fine structure extraction. Application on road crack detection, SPIE-IS&T, vol. 7251 (2009)Google Scholar
- 6.Steckenrider, J.J., Furukawa, T.: Selective pre-processing method for road crack detection with high-speed data acquisition. Int. J. Autom. Eng. (2017)Google Scholar
- 7.Steckenrider, J.J., Furukawa, T.: Detection and classification of stochastic features using a multi-Bayesian approach. In: IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI) (2017)Google Scholar
- 9.Ren, X., Malik, J.: Learning a classification model for segmentation. In: Proceedings Ninth IEEE International Conference on Computer Vision, Nice, France, vol. 1, pp. 10–17 (2003)Google Scholar
- 11.Yao, J., Boben, M., Fidler, S., Urtasun, R.: Real-time coarse-to-fine topologically preserving segmentation. In: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, pp. 2947–2955 (2015)Google Scholar
- 13.Liu, M.Y., Tuzel, O., Ramalingam, S., Chellappa, R.: Entropy rate superpixel segmentation. In: CVPR 2011, Providence, RI, pp. 2097–2104 (2011)Google Scholar
- 15.Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., Süsstrunk, S.: SLIC superpixels, pp. 15–29. EPFL (2010)Google Scholar
- 16.Fernandez, J.: Image processing to detect worms. Uppsala Universitet, Uppsala, Sweden (2010)Google Scholar
- 17.Steckenrider, J.J.: Multi-Bayesian approach to stochastic feature recognition in the context of road crack detection and classification. Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA (2017)Google Scholar