Detecting Pit Defects on Rail Surface Using A Fast Detection Algorithm Based on Relative Gray Value
- Wendi Weng, Houjin Chen
- Corresponding Author
- Wendi Weng
Available Online March 2015.
- https://doi.org/10.2991/iset-15.2015.19How to use a DOI?
- Defect on rail surface, Linear mean filtering, Relative gray value, Fast algorithm
- It is a challenge to detect pit defects on rail surface quickly and accurately in machine vision system. As the pit defects appear randomly, vary in size, distribute discontinuously, and are affected by rust, white noise, shadow and illumination during imaging, pit defects detection has become a difficulty in machine vision field. In this paper, we present a fast detection algorithm based on relative gray value to achieve the requirement of detecting defects on rail surface quickly and accurately. This algorithm uses 1×N dimensional linear mean filtering to improve the detection efficiency. The influence from rust, white noise and environmental impacts are excluded with a set of preprocessing, including offset, contrast of gray values, and image enhancement. Detection accuracy is further improved with Otsu’s binary segmentation method. Experimental results show that this algorithm can detect defects on rail surface quickly and accurately.
- Open Access
- This is an open access article distributed under the CC BY-NC license.
Cite this article
TY - CONF AU - Wendi Weng AU - Houjin Chen PY - 2015/03 DA - 2015/03 TI - Detecting Pit Defects on Rail Surface Using A Fast Detection Algorithm Based on Relative Gray Value BT - First International Conference on Information Science and Electronic Technology (ISET 2015) PB - Atlantis Press UR - https://doi.org/10.2991/iset-15.2015.19 DO - https://doi.org/10.2991/iset-15.2015.19 ID - Weng2015/03 ER -