-
为了克服红外与可见光图像融合时噪声干扰及易产生伪影导致目标轮廓不鲜明、对比度低的缺点，提出一种基于深度模型分割的图像融合方法. 首先，采用深度玻尔兹曼机学习红外与可见光的目标和背景轮廓先验，构建轮廓的深度分割模型，通过Split Bregman迭代算法获取最优能量分割后的红外与可见光图像轮廓；然后再使用非下采样轮廓波变换对源图像进行分解，并针对所分割的背景轮廓采用结构相似度的规则进行系数组合；最后进行非下采样轮廓波反变换重构出融合图像. 数值试验证明，该算法可以有效获取目标和背景轮廓均清晰的融合图像，融合结果不但具有较高的对比度，还能抑制噪声影响，具有有效性.In the infrared and visible light image fusion, the noise interference always exists. There is also the disadvantage that image fusion is easy to produce artifacts which cause blurred edge and low contrast. In order to solve these problems, in this study we propose an image fusion method based on deep model segmentation. First of all, deep Bolzmann machine is adopted to learn prior target and background contour and construct a contour deep segmentation model. After the optimal energy segmentation, Split Bregman iteration is used to obtain the infrared and visible image contour. Then non-subsampled contourlet transform is adopted to decompose the source images. The segmented background contour coefficients are fused by the structure similarity rule. Finally, the fused image is reconstructed by the non-subsampled contourlet inverse transform. The experimental results show that this algorithm can effectively obtain fused images with clear target contour and background contour. The fused images also have high contrast and low noise. The results show that it is an effective method of achieving the infrared and visible image fusion.
[1] Zhang C, Bai L F, Zhang Y 2007 Acta Phys. Sin. 56 3227(in Chinese)[张闯, 柏连发, 张毅 2007 56 3227]
[2] Zhao L Y, Ma Q L, Li X R 2012 Acta Phys. Sin. 61 194204(in Chinese)[赵辽英, 马启良, 厉小润 2012 61 194204]
[3] Ma J F, Hou K, Bao S L, Chen C 2011 Chin. Phys. B 20 028701
[4] Wang X Y, Wang Y X, Yun J J 2011 Chin. Phys. B 20 104202
[5] Zheng H, Zheng C, Yan X S 2012 Chin. J. Sci. Instrum. 33 1613(in Chinese)[郑虹, 郑晨, 闫秀生 2012 仪器仪表学报 33 1613]
[6] Wang J, Peng J Y, He G Q 2013 Acta Armament. 34 815(in Chinese)[王珺, 彭进业, 何贵青 2013 兵工学报 34 815]
[7] Gan T, Feng S T, Nie S P 2011 Acta Phys. Sin. 60 114205(in Chinese)[甘甜, 冯少彤, 聂守平 2011 60 114205]
[8] Kong W, Lei Y 2011 IET Signal Processing 5 75
[9] Shen Y, Dang J W, Feng X 2013 Spectroscopy and Spectral Analysis 33 1506(in Chinese)[沈瑜, 党建武, 冯鑫 2013 光谱学与光谱分析 33 1506]
[10] Mohamed A, Dahl G, Hinton G 2011 IEEE Trans. Audio, Speech, and Language Process. 12 134
[11] Tang Y 2010 In NIPS Workshop on Transfer Learning by Learning Rich Generative Models Vancouver, B C, Canada, December 6, 2010 p2202
[12] Eslami S M, Heess N, Winn J 2012 Computer Vision and Pattern Recognition Rhode Island Provine, America, 16 June, 2012 p983
[13] Wang Z, Bovik A C 2004 IEEE Trans. Image Process. 13 600
[14] Goldstein T, Bresson X, Osher S 2010 J. Sci. Comput. 45 272
[15] Latech L, Lakamper R, Eckhardt U 2000 IEEE Conference on Computer Vision and Pattern Recognition Hilton Head, June 13, USA, 2000 p424
[16] da Cunha A L, Zhou J P, Do M N 2006 IEEE Trans. Image Process. 15 3089
[17] Le Q V, Ngiam J, Coates A, Ng A Y 2011 28th Int. Conf. Machine Learning Bellevue Washington, USA, June 28, 2011 p1209
[18] Zhang Q, Guo B L 2007 J. Infrared Millim. Waves 26 185 (in Chinese)[张强, 郭宝龙 2007 红外与毫米波学报 26 185]
[19] Li X, Qin Y 2011 IET Image Process 5 141
-
[1] Zhang C, Bai L F, Zhang Y 2007 Acta Phys. Sin. 56 3227(in Chinese)[张闯, 柏连发, 张毅 2007 56 3227]
[2] Zhao L Y, Ma Q L, Li X R 2012 Acta Phys. Sin. 61 194204(in Chinese)[赵辽英, 马启良, 厉小润 2012 61 194204]
[3] Ma J F, Hou K, Bao S L, Chen C 2011 Chin. Phys. B 20 028701
[4] Wang X Y, Wang Y X, Yun J J 2011 Chin. Phys. B 20 104202
[5] Zheng H, Zheng C, Yan X S 2012 Chin. J. Sci. Instrum. 33 1613(in Chinese)[郑虹, 郑晨, 闫秀生 2012 仪器仪表学报 33 1613]
[6] Wang J, Peng J Y, He G Q 2013 Acta Armament. 34 815(in Chinese)[王珺, 彭进业, 何贵青 2013 兵工学报 34 815]
[7] Gan T, Feng S T, Nie S P 2011 Acta Phys. Sin. 60 114205(in Chinese)[甘甜, 冯少彤, 聂守平 2011 60 114205]
[8] Kong W, Lei Y 2011 IET Signal Processing 5 75
[9] Shen Y, Dang J W, Feng X 2013 Spectroscopy and Spectral Analysis 33 1506(in Chinese)[沈瑜, 党建武, 冯鑫 2013 光谱学与光谱分析 33 1506]
[10] Mohamed A, Dahl G, Hinton G 2011 IEEE Trans. Audio, Speech, and Language Process. 12 134
[11] Tang Y 2010 In NIPS Workshop on Transfer Learning by Learning Rich Generative Models Vancouver, B C, Canada, December 6, 2010 p2202
[12] Eslami S M, Heess N, Winn J 2012 Computer Vision and Pattern Recognition Rhode Island Provine, America, 16 June, 2012 p983
[13] Wang Z, Bovik A C 2004 IEEE Trans. Image Process. 13 600
[14] Goldstein T, Bresson X, Osher S 2010 J. Sci. Comput. 45 272
[15] Latech L, Lakamper R, Eckhardt U 2000 IEEE Conference on Computer Vision and Pattern Recognition Hilton Head, June 13, USA, 2000 p424
[16] da Cunha A L, Zhou J P, Do M N 2006 IEEE Trans. Image Process. 15 3089
[17] Le Q V, Ngiam J, Coates A, Ng A Y 2011 28th Int. Conf. Machine Learning Bellevue Washington, USA, June 28, 2011 p1209
[18] Zhang Q, Guo B L 2007 J. Infrared Millim. Waves 26 185 (in Chinese)[张强, 郭宝龙 2007 红外与毫米波学报 26 185]
[19] Li X, Qin Y 2011 IET Image Process 5 141
计量
- 文章访问数: 6636
- PDF下载量: 6646
- 被引次数: 0
下载:
