Image Segmentation Algorithm Combining Non-Local Information Interception Kernel Possibilistic Clustering

Jiulun Fan; Yang Yan; Haiyan Yu; Dan Liang; Mengfei Gao

doi:10.3788/LOP57.181010

Journals >Laser & Optoelectronics Progress >Volume 57 >Issue 18 >Page 181010 > Article

Laser & Optoelectronics Progress
Vol. 57, Issue 18, 181010 (2020)

Image Segmentation Algorithm Combining Non-Local Information Interception Kernel Possibilistic Clustering

Jiulun Fan^*, Yang Yan^**, Haiyan Yu^***, Dan Liang, and Mengfei Gao

Author Affiliations

School of Communication and Information Engineering, Xi'an University of Post & Telecommunications, Xi'an, Shaanxi 710121, China

show less

DOI: 10.3788/LOP57.181010 Cite this Article Set citation alerts

Jiulun Fan, Yang Yan, Haiyan Yu, Dan Liang, Mengfei Gao. Image Segmentation Algorithm Combining Non-Local Information Interception Kernel Possibilistic Clustering[J]. Laser & Optoelectronics Progress, 2020, 57(18): 181010 Copy Citation Text

show less

Fig. 1. Images of add different noises. (a) Original image; (b) salt and pepper noise; (c) Gaussian noise; (d) mixed noise

Download full size

Fig. 2. Segmentation results of different algorithms. (a) Salt and pepper noise; (b) Gaussian noise; (c) mixed noise

Download full size

Fig. 3. Image #42049. (a) Original image; (b) salt and pepper noise; (c) Gaussian noise; (d) mixed noise

Download full size

Fig. 4. Segmentation results of different algorithms (#42049). (a) Salt and pepper noise; (b) Gaussian noise; (c) mixed noise

Download full size

Fig. 5. Image Cameraman. (a) Original image; (b) salt and pepper noise; (c) Gaussian noise; (d) mixed noise

Download full size

Fig. 6. Segmentation results of different algorithms (Cameraman). (a) Salt and pepper noise; (b) Gaussian noise; (c) mixed noise

Download full size

Algorithm		PCM	PFCM	FLICM	KFCM_S1	KFCM_S2	KGFCM_S1	KGFCM_S2	Ourproposed
	(0.1)	94.98	95.02	99.71	98.12	96.05	96.65	95.23	99.87
Salt andpepper noise	(0.2)	89.87	90.17	99.69	97.11	94.59	93.67	94.99	99.78
	(0.3)	84.97	84.84	99.62	96.24	91.37	92.92	89.02	99.72
	(0,0.1)	74.13	81.06	99.63	99.02	99.54	96.74	99.54	99.93
Gaussiannoise	(0,0.15)	72.36	76.57	99.07	97.21	99.22	95.21	99.21	99.91
	(0,0.2)	70.44	73.49	99.41	97.98	99.20	93.72	99.17	99.78
	(0.05)and (0, 0.05)	76.74	89.80	99.66	99.12	98.74	97.89	98.19	99.99
Mixednoise	(0.1) and (0, 0.06)	75.77	87.68	99.57	97.11	98.01	96.71	95.37	99.96
	(0.1)and (0, 0.1)	74.30	81.01	99.43	97.39	96.23	95.01	93.64	99.91

Table 1. SA of noisy images by 8 algorithmsunit: %

Algorithm	Salt and pepper noise (0.1)		Gaussian noise (0,0.04)		Mixed noise (0.1) and (0,0.1)
Algorithm	PSNR /dB	SA /%	PSNR /dB	SA /%	PSNR /dB	SA /%
PCM	16.3468	80.90	14.4989	77.82	16.3637	81.20
PFCM	16.3578	80.95	14.3299	77.69	16.9145	80.10
FLICM	21.3922	94.32	21.3847	94.95	21.7668	95.80
KFCM_S1	19.5020	89.37	17.4382	86.20	19.4648	93.59
KFCM_S2	21.8313	95.26	15.9287	79.35	21.5821	95.83
KGFCM_S1	15.3854	80.09	14.1942	78.10	14.9007	78.39
KGFCM_S2	21.3392	95.12	14.0853	78.95	21.3243	95.37
Our proposed	22.5352	95.89	23.1778	96.70	22.6325	96.35

Table 2. Segmentation results of different algorithms (#42049)

Algorithm	Salt andpeppernoise (0.2)	Gaussiannoise(0,0.08)	Mixed noise(0.05) and(0,0.05)
PCM	13.6961	11.9940	12.7491
PFCM	13.8442	11.8895	12.7477
FLICM	18.4335	18.0542	18.3297
KFCM_S1	15.5389	15.0475	15.2110
KFCM_S2	14.9181	14.3703	14.6770
KGFCM_S1	17.8045	17.0568	17.6063
KGFCM_S2	18.0942	16.0394	16.8166
Our proposed	18.5441	18.5010	18.5122

Table 3. PSNR of segmentation results by differentalgorithms (Cameraman image)unit: dB

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information