DI Jing, YIN Shijie, MA Shuai, WANG Guodong
（School of Electronic and Information Engineering， Lanzhou Jiaotong University， Lanzhou 730070， China）
Abstract： Aiming at the problems of misjudgment of focus region, unclear textures and artifacts in multi-focus image fusion, a non-subsampled shearlet transform (NSST) fusion method combining superpixels and guided filter is proposed. Firstly, in the high-frequency fusion, a three-fold guided filter is used to identify the focal region and smooth the focal boundary, and the noise immunity of the image is also enhanced by median denoising. Secondly, gamma correction is performed on the low-frequency subbands based on superpixel to improve the contrast of the image and ensure the continuity of textures. Super-pixel fusion can also maintain the consistency of regional chroma. Finally, the fusion result is composed of brightness and chromaticity subgraphs by inverse Lab transformation. In seven groups of experiments, the spatial frequency (SF), standard deviation (SD), edge intensity (EI) and visual information fidelity(VIFF) values of the proposed algorithm are much higher than those of the contrast algorithms, indicating that the proposed algorithm can preserve the information of textures and contours well, and improve the contrast and sharpness of images.
Key words： multi-focus fusion； non-subsampled shearlet transform(NSST)； superpixel； guided filter
References
［1］ DU C G, HU J W, HU P. Remote sensing imagefusion using semi-supervised convolutional neural network. Journal of Electronic Measurement and Instrumentation, 2021, 35(6): 63-70.
［2］ LI Y, ZHAO J L. A novel medical image fusion method using multi-channel pulse coupled neural networks. IEEE Access, 2020, 8: 157572-157586.
［3］ SHEN Y, HUANG C H, HUANG F, et al. Research progress of infrared and visible image fusion technology. Infrared and Laser Engineering, 2021, 50(9): 152-169.
［4］ LI S T, LI C Y, KANG X D. Development status and future prospect of multi-original remote sensing image fusion. Journal of Remote Sensing, 2021, 25(1): 148-166.
［5］ ROCKINGER O, Fechnert T. Pixel-level image fusion: the case of image sequences//Conference on Signal Processing, Sensor Fusion, and Target Recognition VII, Jul. 17, 1998, Orlando, USA. Bellingham: SPIE, 1998: 378-388.
［6］ STRACK J L, CANDS E J, DONOHO D L. The curvelet transform for image denoising. IEEE Transactions on Image Processing, 2002, 11(6): 670-684.
［7］ DO M N, VETTERLI M. The contourlet transform: an efficient directional multiresolution image representation. IEEE Transactions on Image Processing, 2005, 14(12): 2091-2106.
［8］ GUO K, LABATE D. Optimally sparse multi-dimensional representation using shearlets. SIAM Journal onMathematical Analysis, 2007, 39(1): 298-318.
［9］ EASLEY G R, LABATE D, LIM W. Sparse directional image representations using the discrete shearlet transform. Applied and Computational Harmonic Analysis, 2008, 25(1): 25-46.
［10］ YANG L S, WANG L, GUO Q. Multi-focus image fusion method based on NSST and adaptive PCNN. Computer Science, 2018, 45(12): 217-222.
［11］ JIAO J, WU L D. Morphological filtering and improved PCNN for NSST domain multispectral and panchromatic image fusion. Chinese Journal of Image and Graphics, 2019, 24(3): 435-446.
［12］ LI W, LI Z M. Perception fusion of infrared and visible images in NSST domain. Laser and Optoelectronics Progress, 2021, 58(20): 202-210.
［13］ CHENG F F, FU Z T, HUANG L, et al. Remote sensing image fusion based on adaptive PCNN. Acta Geodaetica et Cartographica Sinica, 2021, 50(10): 1380-1389.
［14］ DING G P, TAO G, LI Y Y, et al. Infrared and visible image fusion based on non-subsampled contourlet transform and guided filter. Acta Armamentarii, 2021, 42(9): 1911-1922.
［15］ LONG Q Y, WANG Z Y, PAN J, et al. Low illumination image enhancement algorithm based on singular value decomposition and guided filtering. Science Technology and Engineering, 2021, 21(12): 5018-5023.
［16］ LI X Y, RAN S Y, LIAN J. Color image segmentation method based on PC-MSPCNN model and SLIC. Progress in Laser & Optoelectronics, 2021, 58(2): 235-242.
［17］ ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2012, 34(11): 2274-2282.
［18］ KALLEL F, HAMIDA A B. A new adaptive gamma correction based algorithm using DWT-SVD for non-contrast CT image enhancement. IEEE Transactions on NanoBioscience, 2017, 16(8): 666-675.
［19］ GANASALS P, PRASAD A D. Contrast enhanced multi sensor image fusion based on guided image filter and NSST. IEEE Sensors Journal, 2019, 20(2): 939-946.
［20］ LIU S, WANG J, LU Y, et al. Multi-focus image fusion based on adaptive dual-channel spiking cortical model innon-subsampled shearlet domain. IEEE Access, 2019, 7: 56367-56388.
基于超像素和引导波的多聚焦图像融合
邸敬， 尹世杰， 马帅， 王国栋
（兰州交通大学 电子与信息工程学院， 甘肃 兰州 730070）
摘要：针对多聚焦图像融合中存在的聚焦区域误判、 纹理不清晰和伪影等问题， 提出一种结合超像素和引导波的非下采样剪切波变换（Non-subsampled shearlet transform, NSST）融合方法。 首先， 在高频融合中使用三重引导滤波达到识别聚焦区域和光滑聚焦边界的效果， 并通过中值去噪增强图像的抗噪性。 其次， 对基于超像素融合的低频子带进行伽马校正提升图像的对比度， 保证图像纹理的连续性， 超像素融合还可以保持区域色度一致性。 最后， 重构的亮度和色度子图通过逆Lab变换进行融合。 在7组实验中， 本算法的空间频率（Spatial frequency, SF）、 标准差（Standard deviation, SD）、 边缘强度（Edge intensity, EI）和视觉信息保真度（Visual information fidelity, VIF）都远大于对比算法的相应值，表明本算法能很好地保留纹理轮廓信息， 提升图像对比度和清晰度。
关键词：多聚焦融合； 非下采样的切波变换； 超像素； 引导波
引用格式：DI Jing, YIN Shijie, MA Shuai, et al． Multi-focus image fusion based on superpixel and guided filter． Journal of Measurement Science and Instrumentation， 2023， 14（3）： 290-298. DOI： 10．3969／j．issn．1674-8042．2023．03．005
[full text view]
