CAJ | 학술논문

在路面裂缝检测中，检测的实时性和准确性同样重要，利用CCD摄像机采集到的视频图像，提出一种基于OpenCV开源平台及改进Canny算子的路面裂缝检测技术。通过对Canny算子的分析与研究，找到其存在的缺陷，利用形态学滤波对原有的滤波方式进行改进，使用Ostu算法实现双阈值的自适应获取。实验结果表明，改进后的Canny算子对路面裂缝检测时，保证了检测的实时性，且提高了检测的精度。
재로면렬봉검측중，검측적실시성화준학성동양중요，이용CCD섭상궤채집도적시빈도상，제출일충기우OpenCV개원평태급개진Canny산자적로면렬봉검측기술。통과대Canny산자적분석여연구，조도기존재적결함，이용형태학려파대원유적려파방식진행개진，사용Ostu산법실현쌍역치적자괄응획취。실험결과표명，개진후적Canny산자대로면렬봉검측시，보증료검측적실시성，차제고료검측적정도。
In the pavement crack detection ,the real‐time and accuracy of detection are both important ,by using video images taken by the CCD camera ,a kind of pavement crack detection technique based on OpenCV open source platforms and the im‐proved Canny edge detection operator was put forward .By analyzing the Canny operator ,its shortcomings were found out .The original filter was improved by using the adaptive morphological filtering ,and the Ostu algorithm was used to achieve the dual threshold gain .The experimental results show that the improved Canny operator on the pavement crack detection not only gua‐rantees the real‐time detection ,but improves the detection accuracy .