基于Tetrolet变换的彩色水下图像清晰化算法

沈瑜; 党建武; 王阳萍; 王博伟

doi:doi:10.3788/AOS201737.0910002

光学学报, 2017, 37 (9): 0910002, 网络出版: 2018-09-07

基于Tetrolet变换的彩色水下图像清晰化算法下载： 970次

A Color Underwater Image Clearness Algorithm Based on Tetrolet Transform

沈瑜党建武 ^*王阳萍王博伟

作者单位

兰州交通大学电子与信息工程学院, 甘肃兰州 730070

图 & 表

图 1. 本文算法流程图

Fig. 1. Flow chart of the proposed algorithm

下载图片查看原文

图 2. 彩色水下图像映射到lαβ三通道。(a) 原始图像; (b) l通道; (c) α通道; (d) β通道

Fig. 2. Maping underwater color image to lαβ three channels. (a) Original image; (b) l channel; (c) α channel; (d) β channel

下载图片查看原文

图 3. 五种基础四格拼板

Fig. 3. Five basic tetrominoes

下载图片查看原文

图 4. Tetrolet变换结构示意图

Fig. 4. Structure diagram of Tetrolet transformation

下载图片查看原文

图 5. 高频细节对比图。(a) Tetrolet变换; (b) 改进的Tetrolet变换

Fig. 5. Comparison of high-frequency details. (a) Tetrolet transform; (b) improved Tetrolet transform

下载图片查看原文

图 6. 双边滤波器在图像边缘附近的作用效果。(a)包含噪声的图像区域; (b)值域滤波; (c)双边滤波; (d)空间域滤波; (e)空间域与值域滤波组合

Fig. 6. Effects on image edge of bilateral filtering. (a) Image region with noises; (b) value domain filtering; (c) bilateral filtering; (d) spatial domain filtering; e) filtering combination of spatial domain and value domain

下载图片查看原文

图 7. Tetrolet处理前后图像。(a) 原始图像; (b) Tetrolet处理后图像

Fig. 7. Images before and after Tetrolet processing. (a) Original images; (b) after Tetrolet processing

下载图片查看原文

图 8. 实验结果对比。(a) 原始图像; (b) Retinex方法; (c) 三通道方法; (d) 本文方法

Fig. 8. Comparison of experiment results. (a) Original images; (b) Retinex method; (c) three-channel method; (d) proposed method

下载图片查看原文

表 1Image1采用3种处理方法的量化比较

Table1. Quantitative results comparison of three methods in Image1

Image1	S_D	A_vG	K	P_SNR
Original image	21.9478	3.2446	11.4797
Retinex method	59.3020	8.7878	11.7245	18.7463
Three-channel method	64.9892	10.5367	11.0100	22.0152
Proposed method	69.1260	14.0523	10.7489	40.9846

查看原文

表 2Image2采用3种处理方法的量化比较

Table2. Quantitative results comparison of three methods in Image2

Image2	S_D	A_vG	K	P_SNR
Original image	37.4333	3.0789	22.7547
Retinex method	67.7373	5.6715	23.0091	24.4258
Three-channel method	70.8437	6.3833	20.7889	31.5185
Proposed method	75.8796	16.1642	17.5998	43.8117

查看原文

表 3Image3采用3种处理方法的量化比较

Table3. Quantitative results comparison of three methods in Image3

Image3	S_D	A_vG	K	P_SNR
Original image	22.1821	4.5231	33.9549
Retinex method	60.0284	12.1882	33.1598	13.8773
Three-channel method	64.7879	15.7308	32.0912	27.8534
Proposed method	68.6630	15.9291	26.9135	39.2829

查看原文

表 4Image4采用3种处理方法的量化比较

Table4. Quantitative results comparison of three methods in Image4

Image4	S_D	A_vG	K	P_SNR
Original image	30.3121	3.5554	12.9045
Retinex method	58.0390	6.8081	13.1663	19.6965
Three-channel method	70.8671	9.2744	13.0921	39.1319
Proposed method	73.3712	16.3484	13.6548	41.8504

查看原文

表 5Image5采用3种处理方法的量化比较

Table5. Quantitative results comparison of three methods in Image5

Image5	S_D	A_vG	K	P_SNR
Original image	35.2730	1.8218	13.2963
Retinex method	63.5605	3.4165	13.5406	20.0278
Three-channel method	74.8518	4.4201	12.1948	32.1510
Proposed method	55.7041	15.9609	13.2990	41.4206

查看原文

表 6Image6采用3种处理方法的量化比较

Table6. Quantitative results comparison of three methods in Image6

Image6	S_D	A_vG	K	P_SNR
Original image	41.3750	5.0257	15.0149
Retinex method	74.2653	9.0967	14.6459	16.4409
Three-channel method	74.8739	9.5739	13.5683	31.4975
Proposed method	66.9649	20.8972	11.5868	42.6501

查看原文

表 7Image7采用3种处理方法的量化比较

Table7. Quantitative results comparison of three methods in Image7

Image7	S_D	A_vG	K	P_SNR
Original image	34.2729	4.9122	10.0089
Retinex method	63.3974	9.2870	10.0493	11.7240
Three-channel method	74.8534	11.1896	9.4062	36.0456
Proposed method	72.8994	18.3051	9.3474	38.6173

查看原文

表 8Image8采用3种处理方法的量化比较

Table8. Quantitative results comparison of three methods in Image8

Image8	S_D	A_vG	K	P_SNR
Original image	23.2133	4.9122	10.0089
Retinex method	55.7064	13.2975	9.1934	18.8354
Three-channel method	74.7517	17.6959	6.7261	40.1223
Proposed method	59.8865	15.7695	6.5751	44.0395

查看原文

沈瑜, 党建武, 王阳萍, 王博伟. 基于Tetrolet变换的彩色水下图像清晰化算法[J]. 光学学报, 2017, 37(9): 0910002. Yu Shen, Jianwu Dang, Yangping Wang, Bowei Wang. A Color Underwater Image Clearness Algorithm Based on Tetrolet Transform[J]. Acta Optica Sinica, 2017, 37(9): 0910002.

基于Tetrolet变换的彩色水下图像清晰化算法下载： 970次

图 1. 本文算法流程图

Fig. 1. Flow chart of the proposed algorithm

图 2. 彩色水下图像映射到lαβ三通道。(a) 原始图像; (b) l通道; (c) α通道; (d) β通道

Fig. 2. Maping underwater color image to lαβ three channels. (a) Original image; (b) l channel; (c) α channel; (d) β channel

图 3. 五种基础四格拼板

Fig. 3. Five basic tetrominoes

图 4. Tetrolet变换结构示意图

Fig. 4. Structure diagram of Tetrolet transformation

图 5. 高频细节对比图。(a) Tetrolet变换; (b) 改进的Tetrolet变换

Fig. 5. Comparison of high-frequency details. (a) Tetrolet transform; (b) improved Tetrolet transform

图 6. 双边滤波器在图像边缘附近的作用效果。(a)包含噪声的图像区域; (b)值域滤波; (c)双边滤波; (d)空间域滤波; (e)空间域与值域滤波组合

Fig. 6. Effects on image edge of bilateral filtering. (a) Image region with noises; (b) value domain filtering; (c) bilateral filtering; (d) spatial domain filtering; e) filtering combination of spatial domain and value domain

图 7. Tetrolet处理前后图像。(a) 原始图像; (b) Tetrolet处理后图像

Fig. 7. Images before and after Tetrolet processing. (a) Original images; (b) after Tetrolet processing

图 8. 实验结果对比。(a) 原始图像; (b) Retinex方法; (c) 三通道方法; (d) 本文方法

Fig. 8. Comparison of experiment results. (a) Original images; (b) Retinex method; (c) three-channel method; (d) proposed method

表 1Image1采用3种处理方法的量化比较

Table1. Quantitative results comparison of three methods in Image1

表 2Image2采用3种处理方法的量化比较

Table2. Quantitative results comparison of three methods in Image2

表 3Image3采用3种处理方法的量化比较

Table3. Quantitative results comparison of three methods in Image3

表 4Image4采用3种处理方法的量化比较

Table4. Quantitative results comparison of three methods in Image4

表 5Image5采用3种处理方法的量化比较

Table5. Quantitative results comparison of three methods in Image5

表 6Image6采用3种处理方法的量化比较

Table6. Quantitative results comparison of three methods in Image6

表 7Image7采用3种处理方法的量化比较

Table7. Quantitative results comparison of three methods in Image7

表 8Image8采用3种处理方法的量化比较

Table8. Quantitative results comparison of three methods in Image8

关于本站 Cookie 的使用提示

全站搜索

基于Tetrolet变换的彩色水下图像清晰化算法 下载： 970次

图 1. 本文算法流程图

Fig. 1. Flow chart of the proposed algorithm

图 2. 彩色水下图像映射到lαβ三通道。(a) 原始图像; (b) l通道; (c) α通道; (d) β通道

Fig. 2. Maping underwater color image to lαβ three channels. (a) Original image; (b) l channel; (c) α channel; (d) β channel

图 3. 五种基础四格拼板

Fig. 3. Five basic tetrominoes

图 4. Tetrolet变换结构示意图

Fig. 4. Structure diagram of Tetrolet transformation

图 5. 高频细节对比图。(a) Tetrolet变换; (b) 改进的Tetrolet变换

Fig. 5. Comparison of high-frequency details. (a) Tetrolet transform; (b) improved Tetrolet transform

图 6. 双边滤波器在图像边缘附近的作用效果。(a)包含噪声的图像区域; (b)值域滤波; (c)双边滤波; (d)空间域滤波; (e)空间域与值域滤波组合

Fig. 6. Effects on image edge of bilateral filtering. (a) Image region with noises; (b) value domain filtering; (c) bilateral filtering; (d) spatial domain filtering; e) filtering combination of spatial domain and value domain

图 7. Tetrolet处理前后图像。(a) 原始图像; (b) Tetrolet处理后图像

Fig. 7. Images before and after Tetrolet processing. (a) Original images; (b) after Tetrolet processing

图 8. 实验结果对比。(a) 原始图像; (b) Retinex方法; (c) 三通道方法; (d) 本文方法

Fig. 8. Comparison of experiment results. (a) Original images; (b) Retinex method; (c) three-channel method; (d) proposed method

表 1Image1采用3种处理方法的量化比较

Table1. Quantitative results comparison of three methods in Image1

表 2Image2采用3种处理方法的量化比较

Table2. Quantitative results comparison of three methods in Image2

表 3Image3采用3种处理方法的量化比较

Table3. Quantitative results comparison of three methods in Image3

表 4Image4采用3种处理方法的量化比较

Table4. Quantitative results comparison of three methods in Image4

表 5Image5采用3种处理方法的量化比较

Table5. Quantitative results comparison of three methods in Image5

表 6Image6采用3种处理方法的量化比较

Table6. Quantitative results comparison of three methods in Image6

表 7Image7采用3种处理方法的量化比较

Table7. Quantitative results comparison of three methods in Image7

表 8Image8采用3种处理方法的量化比较

Table8. Quantitative results comparison of three methods in Image8

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索

基于Tetrolet变换的彩色水下图像清晰化算法下载： 970次