Abstract
To solve the problems of wheel tread damages of railway wagons, a dynamic detecting system was developed and a new method for locating wheel tread peeling and flat spots was presented. Firstly, the wheel tread gray-image was preprocessed for smoothing denoising. Secondly, image edge detection was implemented with combination of the stationary wavelet transform and Canny algorithm and the tread region of the wheel was found on the basis of the width of the standard wheel. Finally, considering the different features of peeling and flat spots, approximate locating of peeling spots was achieved by block segmentation and the seed filling algorithm, and then precise locating of peeling spots was achieved by the Canny and tracking methods, also the locating method of wheel tread flat spots was proposed on the basis of searching of tread edge lines. Experimental results show that at the 0.5 km/h~10 km/h running speed of railway wagons, the correctness rate of peeling locating comes to 96.7% with 3 mm precision and the correctness rate of flat spot locating comes to 97.8%, thus satisfying the requirements of dynamic detection on site.
| Original language | English |
|---|---|
| Pages (from-to) | 31-36 |
| Number of pages | 6 |
| Journal | Tiedao Xuebao/Journal of the China Railway Society |
| Volume | 31 |
| Issue number | 5 |
| DOIs | |
| State | Published - Oct 2009 |
Keywords
- Dynamic detection
- Flat spot
- Peeling spot
- Railway wagon
- Wheel tread