|
Title |
Character Recognition of Low Resolution CCTV Images of Sewer Inspection
|
|
Authors |
김병철(Byeong-Cheol Kim) ; 최창호(Chang-Ho Choi) ; 손병직(Byung-Jik Son) |
|
DOI |
http://dx.doi.org/10.11112/jksmi.2016.20.5.058 |
|
Keywords |
Image processing ; Character recognition ; Low resolution ; Sewer ; CCTV |
|
Abstract |
Recent frequent occurrence of urban sinkhole serves as a momentum of the periodic inspection of sewer pipelines. Sewer inspection using a CCTV device needs a lot of time and efforts. Many of previous studies which reduce the laborious tasks are mainly interested in the developments of image processing S/W and inspection H/W. However there has been no attempt to find meaningful information from the existing CCTV images stored by the sewer maintenance manager. This study adopts a cross-correlation based image processing method and extracts location data of sewer inspection device from CCTV images. As a result of the analysis of time-location relation, it shows strong correlation between the device's stand times and the sewer damages. In case of using this method to investigate sewer inspection CCTV images, it will save the investigator's efforts and improve the sewer maintenance efficiency and reliability.
|