Citation: | Chen Yuzhang, Ye Ting, Cheng Chaojie, et al. Degradation and optimal recovery of underwater turbulent imaging[J]. Opto-Electronic Engineering, 2018, 45(12): 180233. doi: 10.12086/oee.2018.180233 |
[1] | Gilbert G D, Honey R C. Optical turbulence in the sea[J]. Proceedings of SPIE, 1971, 24: 49-56. doi: 10.1117/12.953476 |
[2] | Hou W L, Lee Z, Weidemann A D. Why does the Secchi disk disappear? An imaging perspective[J]. Optics Express, 2007, 15(6): 2791-2802. doi: 10.1364/OE.15.002791 |
[3] | Hou W L, Gray D J, Weidemann A D, et al. Comparison and validation of point spread models for imaging in natural waters[J]. Optics Express, 2008, 16(13): 9958-9965. doi: 10.1364/OE.16.009958 |
[4] | Hou W L. A simple underwater imaging model[J]. Optics Letters, 2009, 34(17): 2688-2690. doi: 10.1364/OL.34.002688 |
[5] | Hou W L, Jarosz E, Woods S, et al. Impacts of underwater turbulence on acoustical and optical signals and their linkage[J]. Optics Express, 2013, 21(4): 4367-4375. doi: 10.1364/OE.21.004367 |
[6] | Hou W L, Woods S, Jarosz E, et al. Optical turbulence on underwater image degradation in natural environments[J]. Applied Optics, 2012, 51(14): 2678-2686. doi: 10.1364/AO.51.002678 |
[7] | Farwell N, Korotkova O. Intensity and coherence properties of light in oceanic turbulence[J]. Optics Communications, 2012, 285(6): 872-875. doi: 10.1016/j.optcom.2011.10.020 |
[8] | Farwell N H, Korotkova O. Multiple phase-screen simulation of oceanic beam propagation[J]. Proceedings of SPIE, 2014, 9224: 922416. doi: 10.1117/12.2062683 |
[9] | Nootz G, Hou W L, Dalgleish F R, et al. Determination of flow orientation of an optically active turbulent field by means of a single beam[J]. Optics Letters, 2013, 38(13): 2185-2187. doi: 10.1364/OL.38.002185 |
[10] | Nootz G, Jarosz E, Dalgleish F R, et al. Quantification of optical turbulence in the ocean and its effects on beam propagation[J]. Applied Optics, 2016, 55(31): 8813-8820. doi: 10.1364/AO.55.008813 |
[11] | Nootz G, Matt S, Kanaev A, et al. Experimental and numerical study of underwater beam propagation in a Rayleigh-Bénard turbulence tank[J]. Applied Optics, 2017, 56(22): 6065-6072. doi: 10.1364/AO.56.006065 |
[12] | Matt S, Hou W L, Woods S, et al. A novel platform to study the effect of small-scale turbulent density fluctuations on underwater imaging in the ocean[J]. Methods in Oceanography, 2014, 11: 39-58. doi: 10.1016/j.mio.2015.01.001 |
[13] | Matt S, Hou W L, Goode W, et al. Introducing SiTTE: a controlled laboratory setting to study the impact of turbulent fluctuations on light propagation in the underwater environment[J]. Optics Express, 2017, 25(5): 5662-5683. doi: 10.1364/OE.25.005662 |
[14] | 孙立颖, 夏珉, 韩捷飞, 等.湍流环境中水下成像系统的调制传递函数研究[J].光学学报, 2016, 36(8): 0801002. Sun L Y, Xia M, Han J F, et al. Research of modulation transfer function of underwater imaging system in turbulent environment[J]. Acta Optica Sinica, 2016, 36(8): 0801002. |
[15] | 蒲欢, 季小玲.海洋湍流中光学成像相关问题研究[J].光学学报, 2016, 36(10): 1026014. Pu H, Ji X L. Problems of optical imaging in oceanic turbulence[J]. Acta Optica Sinica, 2016, 36(10): 1026014. |
[16] | 王马华, 赵正敏, 王士湖, 等.基于改进湍流模型和偏振成像技术的水下退化图像复原方法[J].农业工程学报, 2013, 29(S1): 203-209. Wang M H, Zhao Z M, Wang S H, et al. Restoring method for underwater degraded images based on improved turbulence model and polarization imaging[J]. Transactions of the Chinese Society of Agricultural Engineering, 2013, 29(S1): 203-209. |
[17] | 杨爱萍, 张莉云, 曲畅, 等.基于加权L1正则化的水下图像清晰化算法[J].电子与信息学报, 2017, 39(3): 626-633. Yang A P, Zhang L Y, Qu C, et al. Underwater images visibility improving algorithm with weighted L1 regularization[J]. Journal of Electronics & Information Technology, 2017, 39(3): 626-633. |
[18] | 杨爱萍, 曲畅, 王建, 等.基于水下成像模型的图像清晰化算法[J].电子与信息学报, 2018, 40(2): 298-305. Yang A P, Qu C, Wang J, et al. Underwater image visibility restoration based on underwater imaging model[J]. Journal of Electronics & Information Technology, 2018, 40(2): 298-305. |
[19] | 靳晓娟, 邓志良.基于L1范数和正交梯度算子的超分辨率重建[J].应用光学, 2012, 33(2): 305-312. Jin X J, Deng Z L. Super resolution reconstruction based on L1-norm and orthogonal gradient operator[J]. Journal of Applied Optics, 2012, 33(2): 305-312. |
[20] | Chen Y Z, Yang W L, Tan H Y, et al. Image enhancement for LD based imaging in turbid water[J]. Optik - International Journal for Light and Electron Optics, 2016, 127(2): 517-521. doi: 10.1016/j.ijleo.2015.10.161 |
Overview: At present, the related researches of underwater turbulence mainly include three major categories. The first category is the theoretical calculation based on turbulent structure function and scattering properties. The second category is to use the refractive index power spectra to construct experimental systems for indoor or outdoor experimental measurements and analysis. The third category is the simulation testing based on the PIV system. The existing three categories of research methods have different emphases, and few literatures compare them.
In order to comprehensively and objectively study the degenerate factors of underwater turbulent imaging and optimize the corresponding image restoration algorithms, a reusable submarine imaging experiment system with a turbulent flow controllable condition is established. The circulating water pump is used to provide water force, and the water valve is used to control the turbulent flow field in the laboratory tank. The bubble generator is used to generate micro bubbles, and the bubbles is used as tracer particles. Image sensor is used to obtain the images of sinusoidal stripe target plates under different conditions. In order to reduce the experimental error, the experiment is conducted in a dark environment.
Through indoor and outdoor field experiments, the effect of turbulent flow field, path radiation and fluid media on submarine imaging in turbulent flow are studied. The modulation transfer function (MTF) of the underwater turbulence degradation is correspondingly extracted and analyzed for the three models. The adaptation performance and advantages of the three MTFs are compared and analyzed by using several typical image processing algorithms in turbulent image restoration and super-resolution reconstruction. The objective evaluation criteria such as information capacity (IC), blur metric (BM), and gray mean grads (GMG) are used to compare the objective effect of image processing. The experimental results show that the turbulent flow field, path radiation and fluid media affect the underwater imaging process in turbulent water. The turbulent flow field is the main factor that causes the degradation of underwater imaging in the low spatial frequency. The path radiation and fluid media are the main causes of image degradation. In the restoration of the underwater turbulent degraded image, the MTF of the turbulent flow field is suitable for the image restoration, and the MTFs of the path radiation and turbulent fluid media are suitable for the image super-resolution reconstruction. Compared with other methods of the same class, the information capacity and the gray mean grads value are larger and the blur metric value is smaller.
Stereoscopic structure diagram of the experimental system
Physical map of underwater turbulence system
MTF comparison of experimental system
Blind restoration images of different MTFs. (a) Images of different spatial frequencies; (b) Blind restoration images of FMTF-motion; (c) Blind restoration images of FMTF-path; (d) Blind restoration images of FMTF-tur
Reconstruction results of underwater turbulence stripe images. (a) Stripe image of underwater turbulence; (b) LS method of FMTF-motion; (c) LS method of FMTF-path; (d) LS method of FMTF-tur; (e) BD method of FMTF-motion; (f) BD method of FMTF-path; (g) BD method of FMTF-tur
Super resolution results of underwater turbulence stripe images. (a) Stripe image of underwater turbulence; (b) POCS method of FMTF-motion; (c) POCS method of FMTF-path; (d) POCS method of FMTF-tur; (e) L1-TV method of FMTF-motion; (f) L1-TV method of FMTF-path; (g) L1-TV method of FMTF-tur
Reconstruction results of sea fishes images. (a) Sea fishes image; (b) LS method of FMTF-motion; (c) LS method of FMTF-path; (d) LS method of FMTF-tur; (e) BD method of FMTF-motion; (f) BD method of FMTF-path; (g) BD method of FMTF-tur
Super resolution results of sea fishes images. (a) Sea fishes image; (b) POCS method of FMTF-motion; (c) POCS method of FMTF-path; (d) POCS method of FMTF-tur; (e) L1-TV method of FMTF-motion; (f) L1-TV method of FMTF-path; (g) L1-TV method of FMTF-tur