光照不均匀钢坯缺陷图像的二值化方法
|
摘要
以钢坯低倍组织缺陷的自动检测为背景,针对光照不均匀钢坯图像,研究相应的光照均衡化和二值化方法。首先利用钢坯图像中间亮两边暗的光照分布规律,提出一种基于抛物线拟合的光照均衡化方法。而后依据钢坯图像直方图的单峰特征,即目标(钢坯缺陷)在背景(钢坯组织)中占比极小,提出一种基于直方图梯度的自适应二值化方法,将单峰直方图中低梯度值区与高梯度值区之间的分界点(拐点)作为图像二值化阈值。所提方法源于对实际工程对象的数据分析,经实验证明,能有效分割钢坯图像中的缺陷目标,满足钢胚低倍组织缺陷自动检测的需求。
Illumination equalization and binary segmentation of billet images are discussed with the auto detection for macrostructure defects of steel billets as the applied background. Considering the character of the non-uniform illumination is brighter in middle and darker on both side of the image, a parabolic fitting method is presented to equalise the image illumination. During the binary procedure, the unimodal distribution character of the histogram, that is the objects(defect) pixels locate in the small gradient value region, while the background(billet) pixels are in the far larger gradient value region, is applied and an adaptive binary threshold is got by searching the point of inflection of the histogram gradient. The proposed methods of illumination processing and binary images come from the data of the actual object. The experiments on the actual billet images show that the proposed illumination and binary method can meet the demand of the correct image segment, thus is benefit to get a simple and valid scheme of auto defect detection of steel billets.
Illumination equalization and binary segmentation of billet images are discussed with the auto detection for macrostructure defects of steel billets as the applied background. Considering the character of the non-uniform illumination is brighter in middle and darker on both side of the image, a parabolic fitting method is presented to equalise the image illumination. During the binary procedure, the unimodal distribution character of the histogram, that is the objects(defect) pixels locate in the small gradient value region, while the background(billet) pixels are in the far larger gradient value region, is applied and an adaptive binary threshold is got by searching the point of inflection of the histogram gradient. The proposed methods of illumination processing and binary images come from the data of the actual object. The experiments on the actual billet images show that the proposed illumination and binary method can meet the demand of the correct image segment, thus is benefit to get a simple and valid scheme of auto defect detection of steel billets.
引文
[1]郑伟栋,唐志军,郝强,等.铸坯酸蚀检验法在板坯生产中的应用[J].连铸,2011,(2):44-46.Zheng W D,Tang Z J,Hao Q,et al.Application of acid etching of steel strand in slab production[J].Continuous Casting,2011,(2):44-46.
[2]梁琳,何卫平,雷蕾,等.光照不均图像增强方法综述[J].计算机应用研究,2010,27(5):1625-1628.Liang L,He W P,Lei L,et al.Survey on enhancement methods for non-uniform illumination image[J].Application Research of Computer,2010,27(5):1625-1628
[3]朱珂汉,杨鸣,蒋金涛.显微镜透射光照不均匀图像校正方法[J].计算机应用,2013,33(S1):190-192.Zhu K H,Yang M,Jiang J T.Correction method for microscope transmittance uneven illumination images[J].Journal of Computer Applications,2013,33(S1):190-192.
[4]刘晓玉,吴冰,苏志祁,等.基于数学形态学的棒材自动计数方法[J].控制工程,2013,20(2):286-288,294.Liu X Y,Wu B,Su Z Q,et al.Method for automatic counting of steel bars based on mathematical morphology[J].Control Engineering of China,2013,20(2):286-288,294.
[5]郑国,孔建益,刘源等.重轨弧面缺陷图像光照不均校正方法研究[J].机械设计与制造,2017,增刊:190-193.Zheng G,Kong J Y,Liu Y,et al.Study on correction methods of illumination unevenness for heavy rail curve defect image[J].Machinery Design&Manufacture,2017,Special issue:190-193
[6]Rafael C.Gonzalez,Richard E.Woods,Steven L.Eddins著.数字图像处理(MATLAB版)(第2版)[M].北京:电子工业出版社.2013.Rafael C.Gonzalez,Richard E.Woods,Steven L.Eddins.Digital Image Processing Using MATLAB(Second Edition)[M].Beijing:Publishing House of Electronics Industry.2013
[7]张晓宇,胡士强.基于同态滤波抑制光照变化的视频分割算法[J].模式识别与人工智能,2013,26(1):99-105.Zhang X Y,Hu S Q.Video segmentation algorithm based on homomorphic filtering inhibiting illumination changes[J].Pattern Recognition and Artificial Intelligence,2013,26(1):99-105.
[8]Land E,McCann J.Lightness and Retinex theory[J].Journal of the optical society of America.1971,61(1):1-11.
[9]曾凡锋,树鹏.Retinex在光照不均文本图像中的研究[J].计算机工程与设计,2017,38(11):3072-3079.Zeng F F,Shu P.Research on Retinex in uneven illumination document images[J].Computer Engineering and Design,2017,38(11):3072-3079.
[10]陈志斌,张超,宋岩,等.灰度拉伸Retinex在大动态范围烟雾图像增强中的应用[J].红外与激光工程,2014,43(9):3146-3150.Chen Z B,Zhang C,Song Y,et al.Application of Retinex with grayscale stretching in large dynamic range smoke image enhancement[J].Infrared and Laser Engineering,2014,43(9):3146-3150.
[11]Otsu,N.,A threshold selection method from gray-level histograms[J].IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1):62-66.
[12]郭佳,刘晓玉,吴冰,付晓薇.一种光照不均匀图像的二值化方法[J].计算机应用与软件,2014,31(3):183-202.Guo J,Liu X Y,Wu B,et al.Binarisation method for images acquired under non-uniform illumination[J].Computer Applications and Software,2014,31(3):183-202.
[13]刘振.多光谱图像获取中光照不均匀性的非线性校正算法[J].中国印刷与包装研究,2013,(1):23-27,32.Liu Z.Non-linear correction algorithm of illumination non-uniformity for multi-spectral image[J].China Printing Materials Market,2013,(1):23-27,32.
[14]金守峰,范荻,田明锐.非结构化道路图像光照不均校正方法[J].计算机与数字工程,2016,44(10):2025-2028,2036.Jin S F,Fan D,Tian M R.A correction method for uneven illumination of unstructured road images[J].Computer&Digital Engineering,2016,44(10):2025-2028,2036.
[15]梁志国,李新良,朱振宇.一种基于二次曲线拟合的冲击峰值计算方法[J].计量学报,2015,36(3):309-312.Liang Z G,Li X L,Zhu Z Y.A novel calculation method for peak of impulse waveforms based on parabola curve-fitting[J].Acta Metrologica Sinica,2015,36(3):309-312.
[16]苑会宁,张浩鹏.基于多项式拟合的皮肤镜图像光照不均去除[J].中国体视学与图像分析,2016,(4):386-392.Yuan H N,Zhang H P.Uneven illumination removal for dermoscopy images based on polynomial fitting[J].Chinese Journal of Stereology and Image Analysis,2016,(4):386-392
[17]王娟,王萍,刘敏复杂光照下DPM图像自适应多阈值分割方法研究[J]计算机工程与应用,2018,54(9):194-200.Wang J,Wang P,Liu M.Adaptive multi-thresholds segmentation of DPM barcode image in complex illumination[J].Computer Engineering and Applications,2018,54(9):194-200.
[18]吴丹,蔡晓东,谢月飞等.基于直方图分析和OTSU算法的文字图像二值化[J].计算机与现代化,2013,(7):117-119,234.Wu D,Cai X D,Xie Y F,et al.Binarization of text image based on histogram analysis and OTSU algorithm[J].Computer and Modernization,2013,(7):117-119,234.
[19]段锁林,朱方,严翔.多窗口图像二值化算法研究[J].计算机工程与应用,2017,53(17):212-216.Duan S L,Zhu F,Yan X.Study of multi-window binarization algorithm for image processing[J].Computer Engineering and Applications,2017,53(17):212-216.
[20]张洁玉.基于图像分块的局部阈值二值化方法[J].计算机应用,2017,37(3):827-831.Zhang J Y.Binarization method with local threshold based on image blocks[J].Journal of Computer Applications,2017,37(3):827-831
[2]梁琳,何卫平,雷蕾,等.光照不均图像增强方法综述[J].计算机应用研究,2010,27(5):1625-1628.Liang L,He W P,Lei L,et al.Survey on enhancement methods for non-uniform illumination image[J].Application Research of Computer,2010,27(5):1625-1628
[3]朱珂汉,杨鸣,蒋金涛.显微镜透射光照不均匀图像校正方法[J].计算机应用,2013,33(S1):190-192.Zhu K H,Yang M,Jiang J T.Correction method for microscope transmittance uneven illumination images[J].Journal of Computer Applications,2013,33(S1):190-192.
[4]刘晓玉,吴冰,苏志祁,等.基于数学形态学的棒材自动计数方法[J].控制工程,2013,20(2):286-288,294.Liu X Y,Wu B,Su Z Q,et al.Method for automatic counting of steel bars based on mathematical morphology[J].Control Engineering of China,2013,20(2):286-288,294.
[5]郑国,孔建益,刘源等.重轨弧面缺陷图像光照不均校正方法研究[J].机械设计与制造,2017,增刊:190-193.Zheng G,Kong J Y,Liu Y,et al.Study on correction methods of illumination unevenness for heavy rail curve defect image[J].Machinery Design&Manufacture,2017,Special issue:190-193
[6]Rafael C.Gonzalez,Richard E.Woods,Steven L.Eddins著.数字图像处理(MATLAB版)(第2版)[M].北京:电子工业出版社.2013.Rafael C.Gonzalez,Richard E.Woods,Steven L.Eddins.Digital Image Processing Using MATLAB(Second Edition)[M].Beijing:Publishing House of Electronics Industry.2013
[7]张晓宇,胡士强.基于同态滤波抑制光照变化的视频分割算法[J].模式识别与人工智能,2013,26(1):99-105.Zhang X Y,Hu S Q.Video segmentation algorithm based on homomorphic filtering inhibiting illumination changes[J].Pattern Recognition and Artificial Intelligence,2013,26(1):99-105.
[8]Land E,McCann J.Lightness and Retinex theory[J].Journal of the optical society of America.1971,61(1):1-11.
[9]曾凡锋,树鹏.Retinex在光照不均文本图像中的研究[J].计算机工程与设计,2017,38(11):3072-3079.Zeng F F,Shu P.Research on Retinex in uneven illumination document images[J].Computer Engineering and Design,2017,38(11):3072-3079.
[10]陈志斌,张超,宋岩,等.灰度拉伸Retinex在大动态范围烟雾图像增强中的应用[J].红外与激光工程,2014,43(9):3146-3150.Chen Z B,Zhang C,Song Y,et al.Application of Retinex with grayscale stretching in large dynamic range smoke image enhancement[J].Infrared and Laser Engineering,2014,43(9):3146-3150.
[11]Otsu,N.,A threshold selection method from gray-level histograms[J].IEEE Transactions on Systems,Man,and Cybernetics,1979,9(1):62-66.
[12]郭佳,刘晓玉,吴冰,付晓薇.一种光照不均匀图像的二值化方法[J].计算机应用与软件,2014,31(3):183-202.Guo J,Liu X Y,Wu B,et al.Binarisation method for images acquired under non-uniform illumination[J].Computer Applications and Software,2014,31(3):183-202.
[13]刘振.多光谱图像获取中光照不均匀性的非线性校正算法[J].中国印刷与包装研究,2013,(1):23-27,32.Liu Z.Non-linear correction algorithm of illumination non-uniformity for multi-spectral image[J].China Printing Materials Market,2013,(1):23-27,32.
[14]金守峰,范荻,田明锐.非结构化道路图像光照不均校正方法[J].计算机与数字工程,2016,44(10):2025-2028,2036.Jin S F,Fan D,Tian M R.A correction method for uneven illumination of unstructured road images[J].Computer&Digital Engineering,2016,44(10):2025-2028,2036.
[15]梁志国,李新良,朱振宇.一种基于二次曲线拟合的冲击峰值计算方法[J].计量学报,2015,36(3):309-312.Liang Z G,Li X L,Zhu Z Y.A novel calculation method for peak of impulse waveforms based on parabola curve-fitting[J].Acta Metrologica Sinica,2015,36(3):309-312.
[16]苑会宁,张浩鹏.基于多项式拟合的皮肤镜图像光照不均去除[J].中国体视学与图像分析,2016,(4):386-392.Yuan H N,Zhang H P.Uneven illumination removal for dermoscopy images based on polynomial fitting[J].Chinese Journal of Stereology and Image Analysis,2016,(4):386-392
[17]王娟,王萍,刘敏复杂光照下DPM图像自适应多阈值分割方法研究[J]计算机工程与应用,2018,54(9):194-200.Wang J,Wang P,Liu M.Adaptive multi-thresholds segmentation of DPM barcode image in complex illumination[J].Computer Engineering and Applications,2018,54(9):194-200.
[18]吴丹,蔡晓东,谢月飞等.基于直方图分析和OTSU算法的文字图像二值化[J].计算机与现代化,2013,(7):117-119,234.Wu D,Cai X D,Xie Y F,et al.Binarization of text image based on histogram analysis and OTSU algorithm[J].Computer and Modernization,2013,(7):117-119,234.
[19]段锁林,朱方,严翔.多窗口图像二值化算法研究[J].计算机工程与应用,2017,53(17):212-216.Duan S L,Zhu F,Yan X.Study of multi-window binarization algorithm for image processing[J].Computer Engineering and Applications,2017,53(17):212-216.
[20]张洁玉.基于图像分块的局部阈值二值化方法[J].计算机应用,2017,37(3):827-831.Zhang J Y.Binarization method with local threshold based on image blocks[J].Journal of Computer Applications,2017,37(3):827-831