-
摘要:
为了提高油液磨粒图像的分割效果，优化磨粒自动识别工作的重要环节，提出了一种结合分水岭算法及区域相似度合并的油液磨粒图像自适应分割方法。对于待分割图像，首先通过形态学重构和基于形态学的自适应H-minima技术对梯度图像进行修正，利用分水岭算法完成磨粒图像的一次分割；其次提取分水岭分割后同质区域的Lab颜色特征、局部二值模式（LBP）纹理特征作为区域的量化指标，基于Bhattacharyya系数分别计算区域间的颜色、纹理相似度，设计可以实现权重自适应调整的颜色、纹理特征融合规则，以此来获取同质区域的综合相似度矩阵，实现过分割区域的合并；最后基于统一的后处理算法完成磨粒图像的完整分割。选择60幅磨粒图像对所提方法的分割效果进行测试，结果表明，单幅图像的平均分割速度在12 s左右，分割正确率稳定在90%以上，所提方法避免了用户在分割过程中对图像的交互式处理，较好地平衡了分割效率和分割精度，自适应程度明显提高。
-
关键词:
- H-minima技术 /
- 分水岭算法 /
- 颜色直方图 /
- 局部二值模式(LBP)纹理 /
- Bhattacharyya系数 /
- 区域合并
Abstract:In order to improve the segmentation effect of oil wear debris image and optimize the main content of automatic recognition of wear debris, an adaptive segmentation method of oil wear debris image which combines watershed algorithm and regional similarity has been proposed. First, the gradient image was modified by morphological reconstruction and H-minima technology, and the watershed algorithm was then used to segment the image. Second, after watershed, the Lab color feature and local binary patterns (LBP) texture feature of the homogenous region were extracted as their quantitative indicators, and the color similarity and texture similarity between the regions were calculated based on the Bhattacharyya coefficients. In order to merge the over-segmentation region with much accuracy, an efficient feature fusion rule was designed considering the dynamic weight of color and texture factors. Finally, some post-processing methods were taken to complete the segmentation. Sixty images were selected to test the segmentation effect of the proposed method. The results indicate that the average segmentation speed of single image is about 12 seconds, and the segmentation accuracy is more than 90%. This method avoids the interactive processing when segmenting wear debris images, well balances the segmentation efficiency and segmentation accuracy, and significantly improves the adaptation degree of segmentation program.
-
图 3 分水岭分割区域的颜色直方图提取示例
Figure 3. An example of color histogram extraction from watershed segmentation regions
图 4 分水岭分割区域的纹理直方图提取示例
Figure 4. An example of texture histogram extraction from watershed segmentation regions
图 6 区域合并图像的形态学后处理流程
Figure 6. Post-processing flow of region merging image based on morphology
图 9 本文方法与其他方法的分割结果对比
Figure 9. Comparison of segmentation results between proposed method and other methods
图 11 不同方法的分割精度指标对比
Figure 11. Comparison of segmentation accuracy index among different methods
表 1 分割精度评价指标的定义
Table 1. Definition of segmentation accuracy evaluation index
评价指标 表达式 取值范围 RMR RMR=(SRf-SRft)/SRft (-1, 1) EMR EMR=(FDRf-FDRft)/FDRft (-1, 1) RCD RCD=(SRf∩SRft)/SRft [0, 1] 注：SRf表示分割图像目标区域的面积，SRft表示参考图像目标区域的面积，FDRf、FDRft分别表示以FAENA法[15]提取的分割目标区域、参考目标区域的边界分形维数。RMR、EMR指标的取值有正有负，正值表示分割过程中对目标区域存在过分割问题，负值表示目标区域分割不完整，其绝对值的大小表示错误分割的具体程度。 
下载: 导出CSV
-
[1] 金路, 王静秋.基于形态学算法分割铁谱图像研究[J].机械制造与自动化, 2011(3):10-13. doi: 10.3969/j.issn.1671-5276.2011.03.003JIN L, WANG J Q.Technical research on ferrography division of morphology[J].Machine Building & Automation, 2011(3):10-13(in Chinese). doi: 10.3969/j.issn.1671-5276.2011.03.003 [2] 王静秋, 张龙, 王晓雷.融合颜色聚类和分水岭算法的铁谱图像分割[J].中国矿业大学学报, 2013, 42(5):866-872. http://d.old.wanfangdata.com.cn/Periodical/zgkydxxb201305025WANG J Q, ZHANG L, WANG X L.Ferrographic image segmentation by the method combining k-means clustering and watershed algorithm[J].Journal of China University of Mining & Technology, 2013, 42(5):866-872(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/zgkydxxb201305025 [3] 张龙.蚁群算法在铁谱图像处理中的应用研究[D].南京: 南京航空航天大学, 2013: 31-37. http://cdmd.cnki.com.cn/Article/CDMD-10287-1014005868.htmZHANG L.Application research of ant colony algorithm in ferrography image processing[D].Nanjing: University of Aeronautics and Astronautics, 2013: 31-37(in Chinese). http://cdmd.cnki.com.cn/Article/CDMD-10287-1014005868.htm [4] WANG J, YAO P, LIU W, et al.A hybrid method for the segmentation of a ferrograph image using marker-controlled watershed and grey clustering[J].Tribology Transactions, 2016, 59(3):513-521. doi: 10.1080/10402004.2015.1091534 [5] 张云强, 张培林, 王国德.铁谱图像典型磨粒自动提取方法研究[J].计算机工程与应用, 2013, 49(16):175-178. doi: 10.3778/j.issn.1002-8331.1111-0507ZHANG Y Q, ZHANG P L, WANG G D.Research on automatic extraction of wear particles for ferrographic images[J].Computer Engineering and Applications, 2013, 49(16):175-178(in Chinese). doi: 10.3778/j.issn.1002-8331.1111-0507 [6] 温广瑞, 徐斌, 张志芬, 等.基于差商的油液监测铁谱图像自适应分割[J].光学精密工程, 2017, 25(5):1322-1330. http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201705025WEN G R, XU B, ZHANG Z F, et al.Self-adaptive segmentation of oil monitoring ferrographic image based on difference quotient[J].Optics and Precision Engineering, 2017, 25(5):1322-1330(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201705025 [7] WU H K, WU T H, PENG Y P, et al.Watershed-based morphological separation of wear debris chains for on-line ferrograph analysis[J].Tribology Letters, 2014, 53(2):411-420. doi: 10.1007/s11249-013-0280-1 [8] WU T, WU H, DU Y, et al.Imaged wear debris separation for on-line monitoring using gray level and integrated morphological features[J].Wear, 2014, 316(1-2):19-29. doi: 10.1016/j.wear.2014.04.014 [9] SOILLE P.Morphological image analysis principles and applications[M].Berlin:Springer, 1999:123-140. [10] 余旺盛, 侯志强, 王朝英, 等.基于改进滤波和标记提取的分水岭算法[J].电子学报, 2011, 39(4):825-830. http://d.old.wanfangdata.com.cn/Periodical/dianzixb201104015YU W S, HOU Z Q, WANG C Y, et al.Watershed algorithm based on modified filter and marker-extraction[J].Acta Electronica Sinica, 2011, 39(4):825-830(in Chinese). http://d.old.wanfangdata.com.cn/Periodical/dianzixb201104015 [11] ALI H, LALI M I, NAWAZ M Z, et al.Symptom based automated detection of citrus diseases using color histogram and textural descriptors[J].Computers and Electronics in Agriculture, 2017, 138(C):92-104. http://www.sciencedirect.com/science/article/pii/S0168169916309838 [12] BACKES A R.LBP maps for improving fractal based texture classification[J].Neurocomputing, 2017, 266:1-7. doi: 10.1016/j.neucom.2017.05.020 [13] 宋国云, 范彦斌.基于最大类间方差的数字化铁谱图像分割[J].佛山科学技术学院学报(自然科学版), 2007, 25(1):36-39. doi: 10.3969/j.issn.1008-0171.2007.01.009SONG G Y, FAN Y B.The algorithm for the ferrography image segmentation based on maximum variance between clusters[J].Journal of Foshan University (Natural Science Edition), 2007, 25(1):36-39(in Chinese). doi: 10.3969/j.issn.1008-0171.2007.01.009 [14] 邱丽娟, 宣征南, 张兴芳.基于K-means聚类与最大类间方差的磨粒彩色图像分割[J].润滑与密封, 2014, 39(12):101-104. doi: 10.3969/j.issn.0254-0150.2014.12.021QIU L J, XUAN Z N, ZHANG X F.Debris color image segmentation by K-means clustering and Ostu method[J].Lubrication Engineering, 2014, 39(12):101-104(in Chinese). doi: 10.3969/j.issn.0254-0150.2014.12.021 [15] 齐福柱.基于分形维数的磨粒识别方法研究[D].大连: 大连海事大学, 2014: 16-18. http://d.wanfangdata.com.cn/Thesis/Y2583853QI F Z.Based on the fractal dimension of abrasive identification method research[D].Dalian: Dalian Maritime University, 2014: 16-18(in Chinese). http://d.wanfangdata.com.cn/Thesis/Y2583853 -
下载:
点击查看大图
计量
- 文章访问数: 502
- HTML全文浏览量: 40
- PDF下载量: 345
- 被引次数: 0
