改进中值滤波和形态学的油管裂纹检测算法
|
摘要
油管裂纹检测是保证油管安全工作的前提。裂纹检测的难点是在检测到更多真实裂纹的同时尽可能的抑制噪声。因此,提出了一种改进的中值滤波和多角度形态学相结合的油管裂纹检测算法。首先,利用改进的中值滤波对图像进行去噪;然后,对图像进行小波分解,并用小波模极大值法处理高频分量,同时采用改进的形态学对低频分量进行裂纹检测;最后,将高低频裂纹整合,并将检测出的裂纹依据矩形度进行过滤。MATLAB实验结果显示,对比单独使用形态学法或小波模极大值法,上述算法能在较好抑制噪声的同时保证油管裂纹的连贯性,提高了检测效果。
The detection of pipeline crack is the prerequisite for safe work. The difficulty of crack detection is to detect more real cracks and suppress noise effectively. For this purpose,a new pipeline crack detection method combined median filtering and multi-angle morphological is proposed. Firstly,the median filtering algorithm was used to denoise the image. Secondly,the wavelet decomposition was applied to divide the image into two parts by frequency,then the wavelet modulus maxima method and improved morphology method were used to detect crack in high frequency components and low frequency components respectively. Finally,the high and low frequency cracks were integrated,and the crack that have been detected based on rectangle degrees was filtered. The MATLAB simulation results show that,compared with morphology method or wavelet modulus maxima method,this algorithm can suppress noise and ensure the continuity of the pipeline crack with high accuracy of crack detection.
The detection of pipeline crack is the prerequisite for safe work. The difficulty of crack detection is to detect more real cracks and suppress noise effectively. For this purpose,a new pipeline crack detection method combined median filtering and multi-angle morphological is proposed. Firstly,the median filtering algorithm was used to denoise the image. Secondly,the wavelet decomposition was applied to divide the image into two parts by frequency,then the wavelet modulus maxima method and improved morphology method were used to detect crack in high frequency components and low frequency components respectively. Finally,the high and low frequency cracks were integrated,and the crack that have been detected based on rectangle degrees was filtered. The MATLAB simulation results show that,compared with morphology method or wavelet modulus maxima method,this algorithm can suppress noise and ensure the continuity of the pipeline crack with high accuracy of crack detection.
引文
[1]黄楠.改进的数学形态学图像边缘提取算法研究[J].计算机仿真,2012,29(3):288-291.
[2]王蓓,张根耀,李智.阈值去噪联合数学形态学的胃部图像边缘检测[J].计算机仿真,2015,32(3):375-377.
[3]张德丰. MATLAB数字图像处理[M].北京:机械工业出版社,2012.
[4]杨淑莹.图像识别与项目实践[M].北京:电子工业出版社,2014.
[5]李静.基于油管裂纹的数字图像检测算法的研究与应用[D].西南石油大学,2015.
[6]崔丽群,张月,田鑫.融合双阈值和改进形态学的边缘检测[J].计算机工程与应用,2017,53(9):190-194.
[7]范艳军,吴晓红,罗代升.一种改进的基于小波变换的图像边缘检测算法[J].四川大学学报(自然科学版),2012,49(6):1264-1268.
[8]冈萨雷斯.数字图像处理[M].北京:电子工业出版社,2005.
[9]张铮,王艳平,薛桂香.数字图像处理与机器视觉[M].北京:人民邮电出版社,2010.
[10]张宝峰,王明跃,朱均超.一种零件图像亚像素边缘检测算法[J].计算机仿真,2014,31(2):288-292.
[11]白建军,张欣,陈其松.基于自适应滤波的噪声图像边缘检测方法[J].计算机仿真,2012,29(3):334-337.
[12] J Valenca,D Dias-da-Costa. Automatic crack monitoring using photogrammetry and image processing[J]. Measurement,2013,46:433-441.
[13]李仁兴,等.形态学滤波窄裂纹检测与目标识别[J].无损检测,2009,31(10):796-799.
[14]王耀东,等.基于图像处理的地铁隧道裂缝识别算法研究[J].仪器仪表学报,2014,35(7):1489-1496.
[15]刘丽,等.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.
[16]孙文雅,等.基于图像处理的管道裂缝检测[J].制造业自动化,2012,34(1):36-39.
[2]王蓓,张根耀,李智.阈值去噪联合数学形态学的胃部图像边缘检测[J].计算机仿真,2015,32(3):375-377.
[3]张德丰. MATLAB数字图像处理[M].北京:机械工业出版社,2012.
[4]杨淑莹.图像识别与项目实践[M].北京:电子工业出版社,2014.
[5]李静.基于油管裂纹的数字图像检测算法的研究与应用[D].西南石油大学,2015.
[6]崔丽群,张月,田鑫.融合双阈值和改进形态学的边缘检测[J].计算机工程与应用,2017,53(9):190-194.
[7]范艳军,吴晓红,罗代升.一种改进的基于小波变换的图像边缘检测算法[J].四川大学学报(自然科学版),2012,49(6):1264-1268.
[8]冈萨雷斯.数字图像处理[M].北京:电子工业出版社,2005.
[9]张铮,王艳平,薛桂香.数字图像处理与机器视觉[M].北京:人民邮电出版社,2010.
[10]张宝峰,王明跃,朱均超.一种零件图像亚像素边缘检测算法[J].计算机仿真,2014,31(2):288-292.
[11]白建军,张欣,陈其松.基于自适应滤波的噪声图像边缘检测方法[J].计算机仿真,2012,29(3):334-337.
[12] J Valenca,D Dias-da-Costa. Automatic crack monitoring using photogrammetry and image processing[J]. Measurement,2013,46:433-441.
[13]李仁兴,等.形态学滤波窄裂纹检测与目标识别[J].无损检测,2009,31(10):796-799.
[14]王耀东,等.基于图像处理的地铁隧道裂缝识别算法研究[J].仪器仪表学报,2014,35(7):1489-1496.
[15]刘丽,等.图像纹理特征提取方法综述[J].中国图象图形学报,2009,14(4):622-635.
[16]孙文雅,等.基于图像处理的管道裂缝检测[J].制造业自动化,2012,34(1):36-39.