基于灰度共生矩阵的多尺度分块压缩感知算法

李金凤; 赵雨童; 黄纬然; 郭巾男

doi:doi:10.3788/LOP202158.0410002

激光与光电子学进展, 2021, 58 (4): 0410002, 网络出版: 2021-02-08

基于灰度共生矩阵的多尺度分块压缩感知算法下载： 795次

Multi-Scale Block Compressed Sensing Algorithm Based on Gray-Level Co-Occurrence Matrix

李金凤 ^*赵雨童黄纬然郭巾男

作者单位

沈阳化工大学信息工程学院, 辽宁沈阳 110142

图 & 表

图 1. 本算法的流程图

Fig. 1. Flow chart of our algorithm

下载图片查看原文

图 2. GLCM的像素对路线

Fig. 2. Pixel-to-pixel alignment of the GLCM

下载图片查看原文

图 3. SPL算法流程

Fig. 3. Flow chart of the SPL algorithm

下载图片查看原文

图 4. 图像数据集。(a)Mandrill; (b)Goldhill; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h)Lighthouse; (i)Peppers; (j)Airfield; (k)Zelda; (l)Truck; (m)Tank; (n)Kiel; (o)Fruits; (p)Couple; (q)Mountain; (r)Dollar;(s)Woman; (t)Aerial

Fig. 4. Image date set.(a)Mandrill; (b)Goldhil; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h) Lighthouse; (i) Peppers; (j) Airfield; (k) Zelda; (l) Truck; (m) Tank; (n) Kiel; (o) Fruits; (p)Couple; (q)Mountain; (r)Dollar; (s)Woman; (t)Aerial

下载图片查看原文

图 5. SNR为5 dB时Boat图像的重构效果。 (a)MS-BCS-SPL算法; (b)Document算法; (c)ABCS-SPL算法; (d)本算法

Fig. 5. Reconstruction effect of Boat image when SNR is 5 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

下载图片查看原文

图 6. SNR为10 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 6. Reconstruction effect of Boat image when SNR is 10 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

下载图片查看原文

图 7. SNR为15 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 7. Reconstruction effect of Boat image when SNR is 15 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

下载图片查看原文

表 1不同算法重构图像的PSNR

Table1. PSNR of images reconstructed by different algorithms unit: dB

Algorithm		Sampling rate
Algorithm		0.1	0.2	0.3	0.4	0.5
Mandrill	MS-BCS-SPL	21.49	23.00	24.56	25.53	26.47
	Document	22.86	24.25	25.96	27.55	28.75
	ABCS-SPL	20.68	22.22	23.34	24.54	25.85
	Ours	25.37	26.33	27.41	28.63	29.80
Goldhill	MS-BCS-SPL	29.05	31.07	32.80	33.72	34.69
	Document	30.17	30.85	33.38	34.85	36.05
	ABCS-SPL	25.79	27.71	29.26	30.40	33.50
	Ours	30.89	32.38	33.85	35.23	36.82
Barbara	MS-BCS-SPL	23.89	25.13	26.14	27.44	29.02
	Document	25.21	26.79	28.39	29.97	31.57
	ABCS-SPL	20.59	24.51	26.16	27.91	29.43
	Ours	27.84	28.85	29.96	31.09	32.57
Houses	MS-BCS-SPL	21.11	23.37	25.39	26.30	27.32
	Document	21.96	23.93	25.70	26.36	27.36
	ABCS-SPL	17.40	20.39	21.27	22.96	23.52
	Ours	23.70	25.44	26.97	28.68	30.40
Boat	MS-BCS-SPL	27.41	30.19	32.03	33.20	34.26
	Document	28.69	30.06	31.28	32.37	33.16
	ABCS-SPL	22.40	23.77	27.04	28.93	30.31
	Ours	29.25	30.86	32.54	34.25	36.09
Bridge	MS-BCS-SPL	24.14	25.89	27.76	28.55	29.41
	Document	26.33	27.16	27.84	28.49	29.09
	ABCS-SPL	20.46	23.54	24.11	25.34	26.51
	Ours	26.46	27.85	28.96	30.10	31.47
Home	MS-BCS-SPL	26.51	29.14	31.41	33.78	34.98
	Document	29.34	30.77	31.95	33.23	34.31
	ABCS-SPL	20.40	24.36	27.05	29.36	31.03
	Ours	29.57	31.32	34.82	35.61	36.59
Lighthouse	MS-BCS-SPL	22.93	25.62	27.56	28.64	29.81
	Document	26.61	27.94	29.15	30.16	31.01
	ABCS-SPL	18.60	19.91	20.25	21.51	22.80
	Ours	26.91	28.47	30.08	31.96	33.76
Peppers	MS-BCS-SPL	31.31	34.55	35.80	36.84	37.75
	Document	31.64	33.03	34.22	35.53	36.56
	ABCS-SPL	24.01	26.60	29.60	31.46	33.93
	Ours	32.81	34.63	36.07	37.48	39.33
Algorithm		Sampling rate
Algorithm		0.1	0.2	0.3	0.4	0.5
Airfield	MS-BCS-SPL	24.46	26.82	28.62	29.58	30.54
	Document	25.68	26.69	27.72	28.78	29.59
	ABCS-SPL	21.13	22.24	23.45	24.40	25.18
	Ours	25.97	27.49	28.93	30.32	32.06
Zelda	MS-BCS-SPL	35.02	37.41	39.09	40.04	41.01
	Document	34.97	36.14	37.52	38.56	39.99
	ABCS-SPL	24.58	27.23	32.60	35.28	36.71
	Ours	35.47	37.55	39.16	40.10	41.90
Truck	MS-BCS-SPL	30.87	32.90	34.79	35.65	36.62
	Document	33.68	34.80	35.81	37.35	38.89
	ABCS-SPL	25.12	26.92	28.44	29.95	31.14
	Ours	33.84	35.36	36.84	38.25	39.47
Tank	MS-BCS-SPL	30.93	32.46	34.04	34.85	35.73
	Document	33.86	34.52	35.10	36.18	37.26
	ABCS-SPL	26.05	27.00	28.86	30.09	31.03
	Ours	34.09	35.09	36.24	37.22	38.33
Kiel	MS-BCS-SPL	22.81	24.44	25.51	26.73	28.56
	Document	25.64	26.69	27.75	28.54	29.69
	ABCS-SPL	20.49	21.14	21.92	22.61	23.16
	Ours	25.91	27.13	28.55	29.82	31.56
Fruits	MS-BCS-SPL	19.87	22.21	23.62	24.87	26.23
	Document	24.87	25.69	26.75	27.34	28.08
	ABCS-SPL	19.55	21.65	22.98	24.82	26.19
	Ours	25.07	26.53	27.91	29.32	31.11
Couple	MS-BCS-SPL	26.93	29.55	31.52	32.57	33.70
	Document	29.01	30.65	31.77	32.89	33.84
	ABCS-SPL	22.96	24.52	26.29	27.94	29.37
	Ours	29.38	31.27	32.97	34.71	36.46
Mountain	MS-BCS-SPL	20.33	21.93	23.78	24.55	25.43
	Document	23.30	24.07	24.85	25.59	26.47
	ABCS-SPL	20.16	21.27	22.34	23.22	24.41
	Ours	23.50	24.46	25.31	26.40	27.36
Dollar	MS-BCS-SPL	18.82	20.44	21.42	22.14	22.99
	Document	22.99	23.79	24.85	25.51	26.38
	ABCS-SPL	15.13	15.51	15.86	16.14	16.37
	Ours	23.19	24.58	25.60	26.43	28.17
Woman	MS-BCS-SPL	25.10	27.26	29.12	30.38	31.78
	Document	28.62	29.74	31.08	32.48	33.35
	ABCS-SPL	18.06	24.09	25.77	27.35	28.78
	Ours	28.70	31.92	32.86	33.77	35.56
Aerial	MS-BCS-SPL	24.32	27.01	29.51	30.75	31.96
	Document	26.81	28.36	29.28	30.43	31.12
	ABCS-SPL	22.16	22.19	23.70	25.14	26.65
	Ours	27.07	29.26	30.40	32.08	33.80

查看原文

李金凤, 赵雨童, 黄纬然, 郭巾男. 基于灰度共生矩阵的多尺度分块压缩感知算法[J]. 激光与光电子学进展, 2021, 58(4): 0410002. Jinfeng Li, Yutong Zhao, Weiran Huang, Jinnan Guo. Multi-Scale Block Compressed Sensing Algorithm Based on Gray-Level Co-Occurrence Matrix[J]. Laser & Optoelectronics Progress, 2021, 58(4): 0410002.

基于灰度共生矩阵的多尺度分块压缩感知算法下载： 795次

图 1. 本算法的流程图

Fig. 1. Flow chart of our algorithm

图 2. GLCM的像素对路线

Fig. 2. Pixel-to-pixel alignment of the GLCM

图 3. SPL算法流程

Fig. 3. Flow chart of the SPL algorithm

图 4. 图像数据集。(a)Mandrill; (b)Goldhill; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h)Lighthouse; (i)Peppers; (j)Airfield; (k)Zelda; (l)Truck; (m)Tank; (n)Kiel; (o)Fruits; (p)Couple; (q)Mountain; (r)Dollar;(s)Woman; (t)Aerial

Fig. 4. Image date set.(a)Mandrill; (b)Goldhil; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h) Lighthouse; (i) Peppers; (j) Airfield; (k) Zelda; (l) Truck; (m) Tank; (n) Kiel; (o) Fruits; (p)Couple; (q)Mountain; (r)Dollar; (s)Woman; (t)Aerial

图 5. SNR为5 dB时Boat图像的重构效果。 (a)MS-BCS-SPL算法; (b)Document算法; (c)ABCS-SPL算法; (d)本算法

Fig. 5. Reconstruction effect of Boat image when SNR is 5 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

图 6. SNR为10 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 6. Reconstruction effect of Boat image when SNR is 10 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

图 7. SNR为15 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 7. Reconstruction effect of Boat image when SNR is 15 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

表 1不同算法重构图像的PSNR

Table1. PSNR of images reconstructed by different algorithms unit: dB

关于本站 Cookie 的使用提示

全站搜索

基于灰度共生矩阵的多尺度分块压缩感知算法 下载： 795次

图 1. 本算法的流程图

Fig. 1. Flow chart of our algorithm

图 2. GLCM的像素对路线

Fig. 2. Pixel-to-pixel alignment of the GLCM

图 3. SPL算法流程

Fig. 3. Flow chart of the SPL algorithm

图 4. 图像数据集。(a)Mandrill; (b)Goldhill; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h)Lighthouse; (i)Peppers; (j)Airfield; (k)Zelda; (l)Truck; (m)Tank; (n)Kiel; (o)Fruits; (p)Couple; (q)Mountain; (r)Dollar;(s)Woman; (t)Aerial

Fig. 4. Image date set.(a)Mandrill; (b)Goldhil; (c)Barbara; (d)Houses; (e)Boat; (f)Bridge; (g)Home; (h) Lighthouse; (i) Peppers; (j) Airfield; (k) Zelda; (l) Truck; (m) Tank; (n) Kiel; (o) Fruits; (p)Couple; (q)Mountain; (r)Dollar; (s)Woman; (t)Aerial

图 5. SNR为5 dB时Boat图像的重构效果。 (a)MS-BCS-SPL算法; (b)Document算法; (c)ABCS-SPL算法; (d)本算法

Fig. 5. Reconstruction effect of Boat image when SNR is 5 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

图 6. SNR为10 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 6. Reconstruction effect of Boat image when SNR is 10 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

图 7. SNR为15 dB时Boat图像的重构效果。(a)MS-BCS-SPL算法;(b)Document算法;(c)ABCS-SPL算法;(d)本算法

Fig. 7. Reconstruction effect of Boat image when SNR is 15 dB. (a) MS-BCS-SPL algorithm; (b) Document algorithm; (c) ABCS-SPL algorithm; (d)our algorithm

表 1不同算法重构图像的PSNR

Table1. PSNR of images reconstructed by different algorithms unit: dB

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索

基于灰度共生矩阵的多尺度分块压缩感知算法下载： 795次