视觉显著区域和主动轮廓结合的图像分割算法

doi:10.3778/j.issn.1673-9418.2011043

计算机科学与探索 ›› 2022, Vol. 16 ›› Issue (5): 1155-1168.DOI: 10.3778/j.issn.1673-9418.2011043

视觉显著区域和主动轮廓结合的图像分割算法

何亚茹¹^,², 葛洪伟¹^,²^,⁺()

1.江南大学江苏省模式识别与计算机智能工程实验室,江苏无锡 214122
2.江南大学人工智能与计算机学院,江苏无锡 214122

收稿日期:2020-11-13 修回日期:2021-01-20 出版日期:2022-05-01 发布日期:2022-05-19
通讯作者: + E-mail: ghw8601@163.com
作者简介:何亚茹（1994—）,女,河北石家庄人,硕士研究生,CCF学生会员,主要研究方向为计算机视觉。
葛洪伟（1967—）,男,江苏无锡人,博士,教授,博士生导师,主要研究方向为人工智能与模式识别、图像及视频的处理与分析等。
基金资助:
江苏省研究生创新计划项目(KYLX16_0781);江苏高校优势学科建设工程资助项目

Image Segmentation Algorithm Combining Visual Salient Regions and Active Contour

HE Yaru¹^,², GE Hongwei¹^,²^,⁺()

1. Jiangsu Provincial Engineering Laboratory of Pattern Recognition and Computational Intelligence, Jiangnan Univer-sity, Wuxi, Jiangsu 214122, China
2. School of Artificial Intelligence and Computer Science, Jiangnan University, Wuxi, Jiangsu 214122, China

Received:2020-11-13 Revised:2021-01-20 Online:2022-05-01 Published:2022-05-19
About author:HE Yaru, born in 1994, M.S. candidate, student member of CCF. Her research interest is computer vision.
GE Hongwei, born in 1967, Ph.D., professor, Ph.D. supervisor. His research interests include artificial intelligence and pattern recognition, image and video processing and analysis, etc.
Supported by:
Innovation Program for Graduate of Jiangsu Province(KYLX16_0781);Priority Academic Development Program of Jiangsu Higher Education Institutions

摘要/Abstract

摘要：

传统区域主动轮廓模型在分割弱边缘图像时,演化曲线受背景干扰,易陷入局部极值导致演化速度缓慢;且由于局部项仅考虑空间信息,无法更好保留目标边界,影响分割精度。针对上述问题,首先利用改进的显著性检测方法,对待分割图像进行预处理操作,获取目标候选区域,自动设置初始化轮廓曲线,并将获取的目标先验信息与待分割图像中具有最大对比度的位图相结合,设计自适应符号函数,对优化LoG能量项进行加权,以线性方式融合到RSF模型中,增强模型自适应能力;其次设计新的局部灰度测度,与局部核函数相结合,改进局部能量项,提高模型在弱边缘处的敏感程度,准确定位目标边界。实验结果表明,该模型能够自动设置初始化轮廓,并有效保留目标边缘细节,视觉及定量实验结果证明了该模型优于目前一些主流的主动轮廓模型。

关键词: 视觉显著性检测, 自适应符号函数, 局部灰度测度, 主动轮廓模型, 图像分割

Abstract:

When the traditional regional active contour model is used to segment the weak edge image, the evolution curve is subject to background interference, and it is easy to fall into the local extreme value, which leads to slow evolution speed. Moreover, as the local term only considers the spatial information, it cannot better retain the target boundary, which affects the segmentation accuracy. To solve the above problems, firstly, this paper uses the improved saliency detection algorithm to preprocess the original image, obtains the target candidate regions and automatically sets the initial contour curve. In addition, the obtained priori information of the target is combined with the bitmap with the maximum contrast in the image to be segmented. An adaptive symbolic function is designed to weight the optimized LoG (Laplacian of Gaussian) energy terms, in a linear fashion into RSF (region-scalable fitting) model, improving the adaptive ability of the model. Secondly, a new local grayscale measure is proposed, which is combined with local kernel function to improve the local energy term. It can improve the sensitivity of the model at the weak edge, and accurately locate the target boundary. Experimental results show that this model can automatically set the initial contour and effectively retain the target edge details. Visual and quantitative experimental results show that this model is superior to some mainstream active contour models.

Key words: visual saliency detection, adaptive symbolic function, local gray measure, active contour model, image segmentation

中图分类号:

TP391.41

何亚茹, 葛洪伟. 视觉显著区域和主动轮廓结合的图像分割算法[J]. 计算机科学与探索, 2022, 16(5): 1155-1168.

HE Yaru, GE Hongwei. Image Segmentation Algorithm Combining Visual Salient Regions and Active Contour[J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(5): 1155-1168.

图/表 13

图1 对比度测度改进前后效果

Fig.1 Effect of contrast measure before and after improvement

图2 使用不同聚类算法得到的初始化轮廓

Fig.2 Initial contour obtained by different clustering algorithms

图3 目标种子点

Fig.3 Seeds of target

图4 RBHM模型的有效性

Fig.4 Effectiveness of RBHM model

图5 本文模型流程图

Fig.5 Flow chart of proposed model

表1 分割图6、图7示例图像时各模型中 v、 v 1~ v 5的取值

Table 1 Value of v, v 1 to v 5 in each model segmenting sample images on Fig.6 and Fig.7

Image	$v 1$		$v 2 (255 × 255)$		$v 3 (255 × 255)$		$v 4 (255 × 255)$		$v 5 (255 × 255)$		$v (255 × 255)$
Image	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7
$I 1$	3 000.000 0	3 000.000 0	0.002 0	0.001 0	0.000 9	0.001 1	0.001 0	0.001 0	0.005 9	0.005 9	0.001 2	0.000 8
$I 2$	590.000 0	3 000.000 0	0.001 2	0.001 1	0.001 3	0.001 3	0.001 0	0.001 0	0.005 6	0.005 3	0.000 8	0.000 7
$I 3$	4.000 0	4.200 0	0.002 1	0.004 4	0.002 6	0.004 5	0.002 0	0.003 5	0.011 0	0.012 0	0.001 9	0.004 7
$I 4$	1.500 0	5.500 0	0.001 7	0.001 6	0.002 7	0.001 9	0.003 5	0.003 6	0.008 7	0.008 6	0.002 6	0.002 8
$I 5$	1.200 0	15.000 0	0.180 0	0.012 0	0.055 0	0.007 6	0.035 0	0.013 0	0.120 0	0.120 0	0.021 0	0.011 0
$I 6$	4.100 0	0.500 0	0.110 0	0.160 0	0.180 0	0.099 0	0.140 0	0.140 0	0.290 0	0.370 0	0.021 0	0.021 0

表1 分割图6、图7示例图像时各模型中 v、 v 1~ v 5的取值

Table 1 Value of v, v 1 to v 5 in each model segmenting sample images on Fig.6 and Fig.7

Image	$v 1$		$v 2 (255 × 255)$		$v 3 (255 × 255)$		$v 4 (255 × 255)$		$v 5 (255 × 255)$		$v (255 × 255)$
Image	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7	Fig.6	Fig.7
$I 1$	3 000.000 0	3 000.000 0	0.002 0	0.001 0	0.000 9	0.001 1	0.001 0	0.001 0	0.005 9	0.005 9	0.001 2	0.000 8
$I 2$	590.000 0	3 000.000 0	0.001 2	0.001 1	0.001 3	0.001 3	0.001 0	0.001 0	0.005 6	0.005 3	0.000 8	0.000 7
$I 3$	4.000 0	4.200 0	0.002 1	0.004 4	0.002 6	0.004 5	0.002 0	0.003 5	0.011 0	0.012 0	0.001 9	0.004 7
$I 4$	1.500 0	5.500 0	0.001 7	0.001 6	0.002 7	0.001 9	0.003 5	0.003 6	0.008 7	0.008 6	0.002 6	0.002 8
$I 5$	1.200 0	15.000 0	0.180 0	0.012 0	0.055 0	0.007 6	0.035 0	0.013 0	0.120 0	0.120 0	0.021 0	0.011 0
$I 6$	4.100 0	0.500 0	0.110 0	0.160 0	0.180 0	0.099 0	0.140 0	0.140 0	0.290 0	0.370 0	0.021 0	0.021 0

图6 人工初始化分割结果

Fig.6 Segmentation results of manual initialization

图7 自动初始化分割结果

图7 Segmentation results of automatic initialization

表2 比较六种模型使用不同初始化轮廓所需的迭代时间

Table 2 Iteration time required for six models using different initialized contours s

Image	SPF	SPF*	RSF	RSF*	LGIF	LGIF*	ACML	ACML*	LPF	LPF*	RBHM	Ours
$I 1$	3.7	6.2	14.1	3.7	8.4	4.4	4.4	3.9	1.6	1.4	26.6	6.8
$I 2$	5.5	2.3	3.1	1.8	13.0	1.9	3.1	2.2	1.1	1.1	5.8	2.1
$I 3$	4.8	2.2	6.3	1.4	6.3	2.2	7.2	0.8	8.6	1.3	7.2	0.7
$I 4$	13.7	5.3	20.1	8.7	60.1	9.0	24.3	12.6	16.5	11.8	11.7	19.2
$I 5$	25.4	2.6	251.3	2.2	207.6	2.4	215.9	0.8	71.1	13.5	226.4	1.4
$I 6$	10.2	2.5	54.0	29.0	65.6	29.2	47.8	30.1	13.7	13.7	41.4	24.1
均值	10.6	3.5	58.2	7.8	60.2	8.2	50.5	8.4	18.8	7.1	53.2	9.1

表2 比较六种模型使用不同初始化轮廓所需的迭代时间

Table 2 Iteration time required for six models using different initialized contours s

Image	SPF	SPF*	RSF	RSF*	LGIF	LGIF*	ACML	ACML*	LPF	LPF*	RBHM	Ours
$I 1$	3.7	6.2	14.1	3.7	8.4	4.4	4.4	3.9	1.6	1.4	26.6	6.8
$I 2$	5.5	2.3	3.1	1.8	13.0	1.9	3.1	2.2	1.1	1.1	5.8	2.1
$I 3$	4.8	2.2	6.3	1.4	6.3	2.2	7.2	0.8	8.6	1.3	7.2	0.7
$I 4$	13.7	5.3	20.1	8.7	60.1	9.0	24.3	12.6	16.5	11.8	11.7	19.2
$I 5$	25.4	2.6	251.3	2.2	207.6	2.4	215.9	0.8	71.1	13.5	226.4	1.4
$I 6$	10.2	2.5	54.0	29.0	65.6	29.2	47.8	30.1	13.7	13.7	41.4	24.1
均值	10.6	3.5	58.2	7.8	60.2	8.2	50.5	8.4	18.8	7.1	53.2	9.1

表3 分割图8示例图像时各模型中 v、 v 1~ v 5的取值

Table 3 Value of v, v 1 to v 5 in each model segmenting sample images on Fig.8

Image	$v 1$	$v 2 (255 × 255)$	$v 3 (255 × 255)$	$v 4 (255 × 255)$	$v 5 (255 × 255)$	$v (255 × 255)$
A	3 000.000 0	0.019 0	0.014 0	0.029 0	0.057 0	0.002 9
B	1.600 0	0.000 3	0.000 4	0.000 5	0.000 5	0.000 4
C	190.000 0	0.000 4	0.003 1	0.001 4	0.005 5	0.002 6
D	0.500 0	0.002 3	0.000 7	0.008 8	0.009 1	0.003 1
E	0.800 0	0.028 0	0.033 0	0.045 0	0.050 0	0.012 0
F	7.700 0	0.320 0	0.160 0	0.400 0	1.000 0	0.120 0
G	2.500 0	0.220 0	0.280 0	0.210 0	0.140 0	0.089 0
H	1.900 0	0.008 4	0.014 0	0.005 4	0.016 0	0.001 6
I	3.000 0	0.064 0	0.091 0	0.086 0	0.200 0	0.004 9
J	31.000 0	0.045 0	0.044 0	0.021 0	0.090 0	0.001 6

表3 分割图8示例图像时各模型中 v、 v 1~ v 5的取值

Table 3 Value of v, v 1 to v 5 in each model segmenting sample images on Fig.8

Image	$v 1$	$v 2 (255 × 255)$	$v 3 (255 × 255)$	$v 4 (255 × 255)$	$v 5 (255 × 255)$	$v (255 × 255)$
A	3 000.000 0	0.019 0	0.014 0	0.029 0	0.057 0	0.002 9
B	1.600 0	0.000 3	0.000 4	0.000 5	0.000 5	0.000 4
C	190.000 0	0.000 4	0.003 1	0.001 4	0.005 5	0.002 6
D	0.500 0	0.002 3	0.000 7	0.008 8	0.009 1	0.003 1
E	0.800 0	0.028 0	0.033 0	0.045 0	0.050 0	0.012 0
F	7.700 0	0.320 0	0.160 0	0.400 0	1.000 0	0.120 0
G	2.500 0	0.220 0	0.280 0	0.210 0	0.140 0	0.089 0
H	1.900 0	0.008 4	0.014 0	0.005 4	0.016 0	0.001 6
I	3.000 0	0.064 0	0.091 0	0.086 0	0.200 0	0.004 9
J	31.000 0	0.045 0	0.044 0	0.021 0	0.090 0	0.001 6

图8 六种模型分割结果

Fig.8 Segmentation results of six models

表4 六种模型在图8图像上的分割精度

Table 4 Segmentation accuracy of six models on Fig.8

Image	SPF*		RSF*		LGIF*		ACML*		LPF*		Ours
Image	DSC	JSI	DSC	JSI	DSC	JSI	DSC	JSI	DSC	JSI	DSC	JSI
A	0.893 7	0.807 9	0.912 2	0.838 5	0.913 4	0.840 6	0.912 1	0.838 3	0.911 3	0.837 0	0.926 8	0.863 6
B	0.850 9	0.740 5	0.889 3	0.800 6	0.891 4	0.804 0	0.894 3	0.808 9	0.894 4	0.808 9	0.905 9	0.828 0
C	0.889 2	0.800 4	0.945 5	0.896 7	0.945 1	0.896 0	0.942 9	0.892 0	0.942 7	0.891 6	0.959 7	0.922 5
D	0.901 9	0.821 4	0.905 6	0.827 4	0.904 9	0.826 3	0.905 2	0.826 9	0.905 9	0.828 0	0.925 2	0.860 9
E	0.924 5	0.859 6	0.928 0	0.865 6	0.929 1	0.867 6	0.928 4	0.866 4	0.926 6	0.863 2	0.937 5	0.882 4
F	0.926 7	0.863 4	0.935 7	0.879 2	0.934 0	0.876 2	0.935 1	0.878 1	0.942 5	0.891 2	0.967 1	0.936 3
G	0.859 2	0.753 2	0.889 7	0.801 3	0.887 3	0.797 4	0.889 7	0.801 3	0.267 3	0.154 3	0.934 6	0.877 2
H	0.833 5	0.714 5	0.870 0	0.769 9	0.860 5	0.755 1	0.900 4	0.818 8	0.868 6	0.767 8	0.912 5	0.839 1
I	0.925 2	0.860 8	0.938 4	0.883 9	0.935 4	0.878 6	0.938 0	0.883 3	0.915 1	0.843 5	0.943 4	0.892 8
J	0.869 2	0.768 7	0.863 7	0.760 2	0.869 8	0.769 6	0.892 9	0.806 5	0.603 4	0.432 0	0.921 0	0.853 6
均值	0.887 4	0.799 0	0.907 8	0.832 3	0.907 1	0.831 1	0.913 9	0.842 1	0.817 8	0.731 8	0.933 4	0.875 6

表5 六种模型在图8图像上的分割评价（HD）

Table 5 Segmentation evaluation of six models on Fig.8 (HD)

Image	SPF*	RSF*	LGIF*	ACML*	LPF*	Ours
A	19.65	18.97	18.00	18.97	19.42	9.85
B	25.94	14.00	14.00	14.00	4.00	13.00
C	22.56	6.40	6.40	19.03	19.03	5.39
D	10.63	6.32	6.32	6.32	6.32	6.32
E	18.44	8.25	7.81	7.81	8.25	2.83
F	26.00	26.25	26.25	26.25	26.25	5.00
G	10.77	11.18	11.18	11.18	173.59	8.94
H	84.01	83.73	30.15	9.06	30.15	9.22
I	7.00	23.02	34.54	8.94	16.03	5.10
J	60.21	80.31	80.31	85.62	85.38	10.00
均值	28.52	27.84	23.50	20.72	38.84	7.57

参考文献 28

[1]	HAN B, WU Y. Active contours driven by global and local weighted signed pressure force for image segmentation[J]. Pattern Recognition, 2019, 88: 715-728. DOI URL
[2]	SOOMRO S, MUNIR A, CHOI K N, et al. Fuzzy C-means clustering based active contour model driven by edge scaled region information[J]. Expert Systems with Applications, 2019, 120: 387-396. DOI URL
[3]	KASS M, WITKIN A, TERZOPOULOS D, et al. Snakes: active contour models[J]. International Journal of Computer Vision, 1988, 1(4): 321-331. DOI URL
[4]	GAO S B, YANG J. Saliency-seeded localizing region-based active contour for automatic natural object segmentation[C]// Proceedings of the 21st International Conference on Pattern Recognition, Tsukuba, Nov 11-15, 2012. Washington: IEEE Computer Society, 2012: 3644-3647.
[5]	CASELLES V, CATTE F, COLL T, et al. A geometric model for active contours in image processing[J]. Numerische Mathe-matik, 1993, 66(1): 1-31.
[6]	CHAN T F, VESE L A. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2): 266-277. DOI URL
[7]	LI C M, KAO C Y, GORE J C, et al. Minimization of region-scalable fitting energy for image segmentation[J]. IEEE Trans-actions on Image Processing, 2008, 17(10): 1940-1949.
[8]	DING K, XIAO L F, WENG G R. Active contours driven by local prefitting energy for fast image segmentation[J]. Pattern Recognition Letters, 2018, 104: 29-36. DOI URL
[9]	ZHANG K H, ZHANG L, SONG H H, et al. Active contours with selective local or global segmentation: a new formulation and level set method[J]. Image & Vision Computing, 2010, 28(4): 668-676.
[10]	WANG L, LI C M, SUN Q S, et al. Active contours driven by local and global intensity fitting energy with application to brain MR image segmentation[J]. Computerized Medical Imaging & Graphics, 2009, 33(7): 520-531.
[11]	赵杰, 祁永梅, 潘正勇. 融合模糊全局和双核局部信息的活动轮廓模型[J]. 小型微型计算机系统, 2014, 35(3): 663-666.
	ZHAO J, QI Y M, PAN Z Y. Active contour model of combining fuzzy global and dual-core local information[J]. Journal of Chinese Computer Systems, 2014, 35(3): 663-666.
[12]	DING K Y, XIAO L F, WENG G R. Active contours driven by region-scalable fitting and optimized Laplacian of Gaussian energy for image segmentation[J]. Signal Processing, 2017, 134: 224-233. DOI URL
[13]	陶永鹏, 景雨, 顼聪. 随机森林和水平集的椎骨CT分割方法[J]. 小型微型计算机系统, 2019, 40(6): 1336-1339.
	TAO Y P, JING Y, XU C. Ivertebra CT segmentation method based on random forest and Chan Vese[J]. Journal of Chinese Computer Systems, 2019, 40(6): 1336-1339.
[14]	LI B N, CHUI C K, CHANG S, et al. Integrating spatial fuzzy clustering with level set methods for automated medical image segmentation[J]. Computers in Biology & Medicine, 2011, 41(1): 1-10.
[15]	YANG Y Y, WANG R, FENG C. Level set formulation for automatic medical image segmentation based on fuzzy clus-tering[J]. Signal Processing Image Communication, 2020, 87: 115907. DOI URL
[16]	FU H Z, CAO X C, ZHUO W T. Cluster-based co-saliency detection[J]. IEEE Transactions on Image Processing, 2013, 22(10): 3766-3778. DOI URL
[17]	YUAN Y, LI C, KIM J, et al. Reversion correction and regu-larized random walk ranking for saliency detection[J]. IEEE Transactions on Image Processing, 2018, 27(3): 1311-1322. DOI URL
[18]	林喜兰, 陈秀宏, 肖林云. 局部信息熵的快速混合测地区域活动轮廓模型[J]. 计算机科学与探索, 2016, 10(9): 1332-1340.
	LIN X L, CHEN X H, XIAO L Y. Fast hybrid geodesic region active contour model of local entropy[J]. Journal of Frontiers of Computer Science and Technology, 2016, 10(9): 1332-1340.
[19]	LI C M, XU C Y, GUI C F, et al. Distance regularized level set evolution and its application to image segmentation[J]. IEEE Transactions on Image Processing, 2010, 19(12): 3243-3254. DOI URL
[20]	KIMMEL R, BRUCKSTEIN A M. Regularized Laplacian zero crossings as optimal edge integrators[J]. International Journal of Computer Vision, 2003, 53(3): 225-243. DOI URL
[21]	ZHOU L, YANG Z, ZHOU Z, et al. Salient region detection using diffusion process on a two-layer sparse graph[J]. IEEE Transactions on Image Processing, 2017, 26(12): 5882-5894. DOI URL
[22]	ZONG J J, QIUTS, LI W D, et al. Automatic ultrasound image segmentation based on local entropy and active contour model[J]. Computers & Mathematics with Applications, 2019, 78: 929-943. DOI URL
[23]	CHENG Y. Mean shift, mode seeking and clustering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1995, 17(8): 790-799. DOI URL
[24]	FU Y, CHENG J, LI Z L, et al. Saliency cuts: an automatic approach to object segmentation[C]// Proceedings of the 19th International Conference on Pattern Recognition, Tampa, Dec 8-11, 2008. Washington: IEEE Computer Society, 2008: 1-4.
[25]	BAI X F, WANG W J. Saliency-SVM: an automatic approach for image segmentation[J]. Neurocomputing, 2014, 136(20): 243-255. DOI URL
[26]	李社蕾, 黄梦醒. 改进几何活动轮廓模型的水下图像分割算法研究[J]. 小型微型计算机系统, 2019, 40(3): 671-675.
	LI S L, HUANG M X. Research of underwater image segmentation algorithm based on the improved geometric active contour models[J]. Journal of Chinese Computer Systems, 2019, 40(3): 671-675.
[27]	ALPERT S, GALUN M, BRANDT A, et al. Image segmenta-tion by probabilistic bottom-up aggregation and cue integration[C]// Proceedings of the 2007 IEEE Computer Society Confer-ence on Computer Vision and Pattern Recognition, Minnea-polis, Jun 18-23, 2007. Washington: IEEE Computer Society, 2007: 1-8.
[28]	温立民, 巨永锋, 闫茂德. 基于自然统计特征分布的交通图像雾浓度检测[J]. 电子学报, 2017, 45(8): 1888-1895. DOI
	WEN L M, JU Y F, YAN M D. Inspection of fog density for traffic image based on distribution characteristics of natural statistics[J]. Acta Electronica Sinica, 2017, 45(8): 1888-1895.

视觉显著区域和主动轮廓结合的图像分割算法

Image Segmentation Algorithm Combining Visual Salient Regions and Active Contour

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 28

相关文章 14

编辑推荐

Metrics

[1]	谢娟英, 张凯云. XR-MSF-Unet：新冠肺炎肺部CT图像自动分割模型[J]. 计算机科学与探索, 2022, 16(8): 1850-1864.
[2]	宋姝洁, 崔振超, 陈丽萍, 陈向阳. 多特征融合神经网络的眼底血管分割算法[J]. 计算机科学与探索, 2021, 15(12): 2401-2412.
[3]	钱宝鑫，肖志勇，宋威. 改进的卷积神经网络在肺部图像上的分割应用[J]. 计算机科学与探索, 2020, 14(8): 1358-1367.
[4]	刘辰，肖志勇，杜年茂. 改进的卷积神经网络在医学图像分割上的应用[J]. 计算机科学与探索, 2019, 13(9): 1593-1603.
[5]	师冬丽，李锵，关欣. 结合卷积神经网络和模糊系统的脑肿瘤分割[J]. 计算机科学与探索, 2018, 12(4): 608-617.
[6]	张敬茂，沈艳霞. 多尺度单特征谱分割算法[J]. 计算机科学与探索, 2018, 12(3): 442-451.
[7]	韩铖惠，王慧琴，胡燕. 低对比度火焰图像增强和分割算法研究[J]. 计算机科学与探索, 2018, 12(1): 163-170.
[8]	张娜，王慧琴，胡燕. 粗糙集与区域生长的烟雾图像分割算法研究[J]. 计算机科学与探索, 2017, 11(8): 1296-1304.
[9]	姚龙洋，张清华，胡帅鹏，张强. 基于近似集与粒子群的粗糙熵图像分割方法[J]. 计算机科学与探索, 2016, 10(5): 699-708.
[10]	张新明，涂强，尹欣欣. 混合迁移的高效BBO算法及其在图像分割中的应用[J]. 计算机科学与探索, 2016, 10(10): 1459-1468.
[11]	贺建峰，符增，易三莉，相艳，崔锐. 基于遗传算法的灰度-梯度熵多阈值图像分割[J]. 计算机科学与探索, 2015, 9(8): 995-1003.
[12]	蔡强，刘亚奇，曹健，毛典辉，李海生. 基于SLIC与Delaunay图割的交互式图像分割算法[J]. 计算机科学与探索, 2015, 9(4): 482-490.
[13]	刘立群，火久元，王联国，韩俊英. 混合蛙跳细菌觅食的和声搜索算法及图像应用[J]. 计算机科学与探索, 2015, 9(1): 119-128.
[14]	白雪飞，王文剑. 自适应初始轮廓的Chan-Vese模型图像分割方法[J]. 计算机科学与探索, 2013, 7(12): 1115-1124.