There are some disadvantages existing in traditional artificial method of detecting road rut feature, such as consuming time, consuming energy, low accuracy and danger. In order to solve those problems listed above, a new method of detecting features automatically based on the laser image processing technology is proposed. The first step is to start pretreatments such as fuzzy compensating and enhancing the original images. Then adopting an improved watershed serial segmentation method, we can get binary segmentation images. Used multi-level filtering technology combined with statistical methods, the accurate road rut feature can be got. It has been proved by experiments that the detection method has high accuracy and stability.

Date Published: 30 October 2009
PDF: 6 pages
Proc. SPIE 7494, MIPPR 2009: Multispectral Image Acquisition and Processing, 74941M (30 October 2009); doi: 10.1117/12.833155

Published in SPIE Proceedings Vol. 7494:
MIPPR 2009: Multispectral Image Acquisition and Processing
Faxiong Zhang; Faxiong Zhang, Editor(s)