像素域图像超分辨重建快速算法研究
|
摘要
在实际应用中,高分辨图像通常能为计算机视觉和图像分析等应用提供重要线索,因此,提高图像分辨率的研究就显得尤为重要。但由于诸多客观因素的影响,成像系统获取的图像分辨率往往满足不了实际应用要求。随着图像处理技术的广泛应用,人们对高分辨图像的需求越来越大。为此,采用信号处理手段,即图像超分辨重建技术来提高图像分辨率便成为人们普遍关注的热点。
本文对基于空域和像素域的超分辨重建算法进行了深入研究。首先,提出了一种基于噪声估计的空域超分辨图像重建方法。该方法能避免由部分模型估计误差引起的重建结果的明显波动。实验结果表明,本文提出的算法在主、客观方面都获得了较好的效果,在抑制噪声的同时,能进一步提高重建图像的视觉质量。
其次,针对全局图像配准对复杂的局部运动存在的不足,本文提出了一种基于Zernike矩的像素域超分辨图像重建方法。该方法通过对目标图像逐像素的复原,避免了局部移位、旋转等运动对重建过程的干扰,具有重建质量高,噪声鲁棒性强的特点。实验结果表明,本文提出的方法在主观和客观上均具有良好性能。
最后,为使基于Zernike矩的像素域超分辨图像重建方法有效地用于现实应用,本文提出了一种基于分块缝合的像素域超分辨图像快速重建算法,该方法将逐像素的重建变为逐图像块的重建。基于图像缝合技术的思想,极大地加快了重建算法的速度。实验表明该算法在不明显降低重建质量的同时,明显降低了计算复杂度,提高了图像的重建速度。
In practical application,high resolution images usually provide more important clues for computer vision and image analysis applications. Therefore, the research for improving image resolution is particularly important. But with some interference of objective factors, image resolution obtained by imaging system often can't satisfy the actual application requirements. With the wide use of image processing, the demand of high resolution image is growing more and more. For this reason, rhe method based on the signal processing, namely image super resolution reconstruction technology, became widely concerned hot improve to image resolution.
Thesis has deeply researched based on the super resolution reconstruction in airspace and the pixel domain. First of all, it has proposed a super resolution reconstruction method based on noise estimates in airspace. This method avoids the fluctuation of results caused by estimate errors with some models. Experimental results show that the proposed algorithm has good results in the subjective and objective and further raise the image quality with restraining the noise.
Secondly, because insufficient of global image registration with partial motion. This paper has proposed the image reconstruction method based on the Zernike moment in pixel domain. Through recovering the target image per-pixel and avoiding interference of the local shift, and rotated condition in the reconstruction process, this method has high quality and noise robustness. The proposed approach has good performance in subjective and objective.
Finally, in order to make the method called Zernike moment based super resolution reconstruction can be effectively applied to reality. This paper has presents a fast algorithm based on image quilting in pixel domain. This method has accelerated the speed of algorithm with turning the reconstruction per-pixel to per- block and combining the technique of image quilting. Experiments show that the proposed algorithm has greatly reduced the computational complexity with keeping the quality of reconstruction image.
本文对基于空域和像素域的超分辨重建算法进行了深入研究。首先,提出了一种基于噪声估计的空域超分辨图像重建方法。该方法能避免由部分模型估计误差引起的重建结果的明显波动。实验结果表明,本文提出的算法在主、客观方面都获得了较好的效果,在抑制噪声的同时,能进一步提高重建图像的视觉质量。
其次,针对全局图像配准对复杂的局部运动存在的不足,本文提出了一种基于Zernike矩的像素域超分辨图像重建方法。该方法通过对目标图像逐像素的复原,避免了局部移位、旋转等运动对重建过程的干扰,具有重建质量高,噪声鲁棒性强的特点。实验结果表明,本文提出的方法在主观和客观上均具有良好性能。
最后,为使基于Zernike矩的像素域超分辨图像重建方法有效地用于现实应用,本文提出了一种基于分块缝合的像素域超分辨图像快速重建算法,该方法将逐像素的重建变为逐图像块的重建。基于图像缝合技术的思想,极大地加快了重建算法的速度。实验表明该算法在不明显降低重建质量的同时,明显降低了计算复杂度,提高了图像的重建速度。
In practical application,high resolution images usually provide more important clues for computer vision and image analysis applications. Therefore, the research for improving image resolution is particularly important. But with some interference of objective factors, image resolution obtained by imaging system often can't satisfy the actual application requirements. With the wide use of image processing, the demand of high resolution image is growing more and more. For this reason, rhe method based on the signal processing, namely image super resolution reconstruction technology, became widely concerned hot improve to image resolution.
Thesis has deeply researched based on the super resolution reconstruction in airspace and the pixel domain. First of all, it has proposed a super resolution reconstruction method based on noise estimates in airspace. This method avoids the fluctuation of results caused by estimate errors with some models. Experimental results show that the proposed algorithm has good results in the subjective and objective and further raise the image quality with restraining the noise.
Secondly, because insufficient of global image registration with partial motion. This paper has proposed the image reconstruction method based on the Zernike moment in pixel domain. Through recovering the target image per-pixel and avoiding interference of the local shift, and rotated condition in the reconstruction process, this method has high quality and noise robustness. The proposed approach has good performance in subjective and objective.
Finally, in order to make the method called Zernike moment based super resolution reconstruction can be effectively applied to reality. This paper has presents a fast algorithm based on image quilting in pixel domain. This method has accelerated the speed of algorithm with turning the reconstruction per-pixel to per- block and combining the technique of image quilting. Experiments show that the proposed algorithm has greatly reduced the computational complexity with keeping the quality of reconstruction image.
引文
[1] Chaudhuri S. Super–Resolution Imaging. Norwell, MA: Kluwer Academic, 2001.
[2] Park S C, Park M K, Kang M G. Super-resolution image resolution: a technical overview. IEEE Signal Processing Magazine., 2003, vol.20, no. 3, pp. 21-36.
[3] Tsai R Y, Huang T S. A Single-Frame Super-Resolution Innovative Approach. Advances in Artificial Intelligence Lecture Notes in Computer Science., 2007, vol. 4827/2007, pp. 640-649.
[4] Tekalp A M, Ozkan M K, Sezan M I. High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration. Proceedings of IEEE International Conference on Acoustics Speech and Signal Processing. San Francisco: IEEE Press., 1992, pp. 169-172.
[5] Kim S P, Su W Y. Recursive high-resolution reconstruction of blurred multiframe images. IEEE Transactions on Image Processing., 1993, vol. 2, no.4, pp. 534-539.
[6] Rhee S H, Kang M G. Discrete cosine transform based regularized high-resolution image reconstruction algorithm. Optical Engineering., 1999, vol. 38, no. 8, pp. 1348-1356.
[7] Komatsu T, Igarashi T, Aizawa K, et al. Very high resolution imaging scheme with multiple different-aperture cameras. Signal Processing: Image Communication., 1993, vol. 5, pp. 511-526.
[8] Nguyen N, Milanfar P. A wavelet-based interpolation-restoration method for superresolution. Circuits, Systems, and Signal Processing., 2000, vol. 19, no. 4, pp. 321-338.
[9] Chappalli M B, Bose N K. Simultaneous noise filtering and super-resolution with second-generation wavelets. IEEE Signal Processing Letters., 2005, vol. 12, no. 11, pp. 772-775.
[10] Stark H, Oskoui P. High-resolution image recovery from image plane arrays, using convex projections. Journal of the Optical Society of America (Series A)., 1989, vol. 6, no. 11, pp. 1715-1726.
[11] Ozkan M K, Tekalp A M, Sezan M I. POCS-based restoration of space-varying blurred images. IEEE Transactions on Image Processing., 1994, vol. 3, no. 4, pp. 450-454.
[12] Patti A J, Sezan M I, Tekalp A M. Superresolution video reconstruction with arbitrary sampling lattices and nonzero aperture time. IEEE Transactions on Image Processing., 1997, vol. 6, no. 8, pp. 1064-1076.
[13] Eren P E, Sezan M I., Tekalp A M. Robust, object-based high-resolution image reconstruction from low-resolution video. IEEE Transactions on Image Processing., 1997, vol. 6, no. 10, pp. 1446-1451.
[14] Schulz R R, Stevenson R L. Extraction of high-resolution frames from video sequences. IEEE Transactions on Image Processing., 1996, vol. 5, no. 6, pp. 996-1011.
[15] Hardie R C, Barnard K J, Armstrong E E. Joint MAP registration and high-resolution image estimation using a sequence of undersampled images. IEEE Transactions on Image Processing., 1997, vol. 6, no. 12, pp. 1621-1633.
[16] Cheeseman P, Kanefsky B, Kraft R. Super-resolved surface reconstruction from multiple images. NASA Technical Report FIA-94- 12., 1994.
[17] Hunt B R, Sementill P J. Description of a poisson imagery super-resolution algorithm. Astronomical Data Analysis Software and Systems I. San Francisco: Astronomical Society of the Pacific., 1992, vol. 25, pp. 196-199.
[18] Elad M, Feuer A. Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images. IEEE Transactions on Image Processing., 1997, vol. 6, no. 12, pp. 1646-1658.
[19] Irani M, Peleg S. Improving resolution by image registration. CVGIP: Graphical Models and Image Processing., 1991, vol. 53, no. 3, pp. 231-239.
[20] Mann S, Picard R W. Virtual bellows: constructing high quality stills from video. Proceedings of IEEE International Conference on Image Processing. Austin, TX. , 1994, vol. 1, pp. 363-367.
[21] M. Protter, M. Elad, H. Takeda, P. Milanfar. Generalizing the Nonlocal-Means to Super-Resolution Reconstruction. IEEE Trans. Image Process., Jan. 2009, vol. 18, no. 1, pp. 349–366.
[22] Freeman W T, Jones T R, Pasztor E C. Example-based super resolution. IEEE Computer Graphics and Applications., 2002, vol. 22, no. 2, pp. 56-65.
[23] Shechtman E, Caspi Y, Irani M. Space-time super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence., 2005, vol. 27, no. 4, pp. 531-545.
[24] Peled S, Yeshurun Y. Superresolution in MRI: Application to human white matter fiber tract visualization by diffusion tensor imaging. Magnetic Resonancein Medicine., 2001, vol. 45, no. 1, pp. 29-35.
[25] Elad M, Hel-Or Y. A fast super-resolution reconstruction algorithm for pure translational motion and common space-invariant blur. IEEE Transactions on Image Processing., 2001, vol. 10, no. 8, pp. 1187-1193.
[26] M.Elad and A.Feuer. Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images. IEEE Trans. Image Processing., Dec, 1997, vol.6, no.12, pp. 1646-1658.
[27] S.Farsiu, M.D.Robinson, M,.Elad and P.Milanfar. Fast and robust multiframe super resolution. IEEE Trans. Image Processing., 2004, vol.13, no.10, pp.1327-1344.
[28] Xuelong Li, Yanting Hu, Xinbo Gao, et al. A multi-frame image super-resolution method. Signal Processing., Feb. 2010, vol. 90, no. 2, pp. 405-414.
[29] Rafael C.Gonzale, Richard E.Woods, Steven L.Eddins,阮秋琦等译.数字图像处理(Digital Image Processing Using MATLAB).电子工业出版社, Sep. 2005.
[30] A. Buades, B. Coll, and J. M. Morel. A review of image de-noising algorithms, with a new one. Multiscale Model. Simul., 2005, vol. 4, no. 2, pp. 490–530.
[31] C. Tomasi and R. Manduchi. Bilateral filtering for gray and color images. in Proc. IEEE Int. Conf. Computer Vision., Jan. 1998, pp. 836–846.
[32] L. Rudin, S. Osher, and E. Fatemi. Nonlinear total variation based noise removal algorithms. Phys. D., Nov. 1992, vol. 60, no. 1-4, pp. 259–268.
[33] G.Wolberg. Digital Image Warping. Washington. DC:IEEE Computer Soc Press., 1990.
[34] Ken Turkowski. Flters for common resampling tasks. Academic Press Professional., Oct. 1990, pp. 147-165.
[35] M. K. Hu. Visual Pattern Recognition by Moment Invariant. IRE Trans Information Theory., Feb. 1962, vol. 8, no. 2, pp. 179-187.
[36] M. R. Teague. Image Analysis via the General Theory of Moments. J. Opt. Soc. Amer., Aug. 1980, vol. 70, no. 8, pp. 920-930.
[37] A. Khotanzad, Y.H. Hong. Invariant image recognition by Zernike moments. IEEE Trans. Pattern Anal. Mach. Intell., May. 1990, vol. 12, no. 5, pp. 489–497.
[38] S. Farsiu, D. Robinson, M. Elad, and P. Milanfar. Robust shift and add approach to super resolution. in Proc. SPIE Conf. Applications of Digital Signal and Image Processing., Aug. 2003, pp. 121–130.
[39] S. Farsiu, D. Robinson, M. Elad, and P. Milanfar. Fast and robust multiframe superresolution. IEEE Trans. Image Process., Oct. 2004, vol. 13, no. 10, pp. 1327-1344.
[40] X. Gao, W. Lu, D. Tao, and X. Li. Image Quality Assessment Based on Multiscale Geometric Analysis. IEEE Trans. Image Process., 2009, vol. 18, no. 7, pp. 1409–1423.
[41] A.A. Efros, W.T. Freeman. Image quilting for texture synthesis and transfer. in Proc. the ACM Conference on Computer Graphics and Interactive Techniques, Los Angeles, California, USA., Aug. 2001, vol. 12–17, pp. 341–346.
[42] Bing Xiao, Xinbo Gao, Dacheng Tao, et al. Photo-sketch synthesis and recognition based on subspace learning. Neurocomputing., Jan. 2010, vol. 73, no. 4-6, pp. 840-852.
[2] Park S C, Park M K, Kang M G. Super-resolution image resolution: a technical overview. IEEE Signal Processing Magazine., 2003, vol.20, no. 3, pp. 21-36.
[3] Tsai R Y, Huang T S. A Single-Frame Super-Resolution Innovative Approach. Advances in Artificial Intelligence Lecture Notes in Computer Science., 2007, vol. 4827/2007, pp. 640-649.
[4] Tekalp A M, Ozkan M K, Sezan M I. High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration. Proceedings of IEEE International Conference on Acoustics Speech and Signal Processing. San Francisco: IEEE Press., 1992, pp. 169-172.
[5] Kim S P, Su W Y. Recursive high-resolution reconstruction of blurred multiframe images. IEEE Transactions on Image Processing., 1993, vol. 2, no.4, pp. 534-539.
[6] Rhee S H, Kang M G. Discrete cosine transform based regularized high-resolution image reconstruction algorithm. Optical Engineering., 1999, vol. 38, no. 8, pp. 1348-1356.
[7] Komatsu T, Igarashi T, Aizawa K, et al. Very high resolution imaging scheme with multiple different-aperture cameras. Signal Processing: Image Communication., 1993, vol. 5, pp. 511-526.
[8] Nguyen N, Milanfar P. A wavelet-based interpolation-restoration method for superresolution. Circuits, Systems, and Signal Processing., 2000, vol. 19, no. 4, pp. 321-338.
[9] Chappalli M B, Bose N K. Simultaneous noise filtering and super-resolution with second-generation wavelets. IEEE Signal Processing Letters., 2005, vol. 12, no. 11, pp. 772-775.
[10] Stark H, Oskoui P. High-resolution image recovery from image plane arrays, using convex projections. Journal of the Optical Society of America (Series A)., 1989, vol. 6, no. 11, pp. 1715-1726.
[11] Ozkan M K, Tekalp A M, Sezan M I. POCS-based restoration of space-varying blurred images. IEEE Transactions on Image Processing., 1994, vol. 3, no. 4, pp. 450-454.
[12] Patti A J, Sezan M I, Tekalp A M. Superresolution video reconstruction with arbitrary sampling lattices and nonzero aperture time. IEEE Transactions on Image Processing., 1997, vol. 6, no. 8, pp. 1064-1076.
[13] Eren P E, Sezan M I., Tekalp A M. Robust, object-based high-resolution image reconstruction from low-resolution video. IEEE Transactions on Image Processing., 1997, vol. 6, no. 10, pp. 1446-1451.
[14] Schulz R R, Stevenson R L. Extraction of high-resolution frames from video sequences. IEEE Transactions on Image Processing., 1996, vol. 5, no. 6, pp. 996-1011.
[15] Hardie R C, Barnard K J, Armstrong E E. Joint MAP registration and high-resolution image estimation using a sequence of undersampled images. IEEE Transactions on Image Processing., 1997, vol. 6, no. 12, pp. 1621-1633.
[16] Cheeseman P, Kanefsky B, Kraft R. Super-resolved surface reconstruction from multiple images. NASA Technical Report FIA-94- 12., 1994.
[17] Hunt B R, Sementill P J. Description of a poisson imagery super-resolution algorithm. Astronomical Data Analysis Software and Systems I. San Francisco: Astronomical Society of the Pacific., 1992, vol. 25, pp. 196-199.
[18] Elad M, Feuer A. Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images. IEEE Transactions on Image Processing., 1997, vol. 6, no. 12, pp. 1646-1658.
[19] Irani M, Peleg S. Improving resolution by image registration. CVGIP: Graphical Models and Image Processing., 1991, vol. 53, no. 3, pp. 231-239.
[20] Mann S, Picard R W. Virtual bellows: constructing high quality stills from video. Proceedings of IEEE International Conference on Image Processing. Austin, TX. , 1994, vol. 1, pp. 363-367.
[21] M. Protter, M. Elad, H. Takeda, P. Milanfar. Generalizing the Nonlocal-Means to Super-Resolution Reconstruction. IEEE Trans. Image Process., Jan. 2009, vol. 18, no. 1, pp. 349–366.
[22] Freeman W T, Jones T R, Pasztor E C. Example-based super resolution. IEEE Computer Graphics and Applications., 2002, vol. 22, no. 2, pp. 56-65.
[23] Shechtman E, Caspi Y, Irani M. Space-time super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence., 2005, vol. 27, no. 4, pp. 531-545.
[24] Peled S, Yeshurun Y. Superresolution in MRI: Application to human white matter fiber tract visualization by diffusion tensor imaging. Magnetic Resonancein Medicine., 2001, vol. 45, no. 1, pp. 29-35.
[25] Elad M, Hel-Or Y. A fast super-resolution reconstruction algorithm for pure translational motion and common space-invariant blur. IEEE Transactions on Image Processing., 2001, vol. 10, no. 8, pp. 1187-1193.
[26] M.Elad and A.Feuer. Restoration of a single superresolution image from several blurred, noisy, and undersampled measured images. IEEE Trans. Image Processing., Dec, 1997, vol.6, no.12, pp. 1646-1658.
[27] S.Farsiu, M.D.Robinson, M,.Elad and P.Milanfar. Fast and robust multiframe super resolution. IEEE Trans. Image Processing., 2004, vol.13, no.10, pp.1327-1344.
[28] Xuelong Li, Yanting Hu, Xinbo Gao, et al. A multi-frame image super-resolution method. Signal Processing., Feb. 2010, vol. 90, no. 2, pp. 405-414.
[29] Rafael C.Gonzale, Richard E.Woods, Steven L.Eddins,阮秋琦等译.数字图像处理(Digital Image Processing Using MATLAB).电子工业出版社, Sep. 2005.
[30] A. Buades, B. Coll, and J. M. Morel. A review of image de-noising algorithms, with a new one. Multiscale Model. Simul., 2005, vol. 4, no. 2, pp. 490–530.
[31] C. Tomasi and R. Manduchi. Bilateral filtering for gray and color images. in Proc. IEEE Int. Conf. Computer Vision., Jan. 1998, pp. 836–846.
[32] L. Rudin, S. Osher, and E. Fatemi. Nonlinear total variation based noise removal algorithms. Phys. D., Nov. 1992, vol. 60, no. 1-4, pp. 259–268.
[33] G.Wolberg. Digital Image Warping. Washington. DC:IEEE Computer Soc Press., 1990.
[34] Ken Turkowski. Flters for common resampling tasks. Academic Press Professional., Oct. 1990, pp. 147-165.
[35] M. K. Hu. Visual Pattern Recognition by Moment Invariant. IRE Trans Information Theory., Feb. 1962, vol. 8, no. 2, pp. 179-187.
[36] M. R. Teague. Image Analysis via the General Theory of Moments. J. Opt. Soc. Amer., Aug. 1980, vol. 70, no. 8, pp. 920-930.
[37] A. Khotanzad, Y.H. Hong. Invariant image recognition by Zernike moments. IEEE Trans. Pattern Anal. Mach. Intell., May. 1990, vol. 12, no. 5, pp. 489–497.
[38] S. Farsiu, D. Robinson, M. Elad, and P. Milanfar. Robust shift and add approach to super resolution. in Proc. SPIE Conf. Applications of Digital Signal and Image Processing., Aug. 2003, pp. 121–130.
[39] S. Farsiu, D. Robinson, M. Elad, and P. Milanfar. Fast and robust multiframe superresolution. IEEE Trans. Image Process., Oct. 2004, vol. 13, no. 10, pp. 1327-1344.
[40] X. Gao, W. Lu, D. Tao, and X. Li. Image Quality Assessment Based on Multiscale Geometric Analysis. IEEE Trans. Image Process., 2009, vol. 18, no. 7, pp. 1409–1423.
[41] A.A. Efros, W.T. Freeman. Image quilting for texture synthesis and transfer. in Proc. the ACM Conference on Computer Graphics and Interactive Techniques, Los Angeles, California, USA., Aug. 2001, vol. 12–17, pp. 341–346.
[42] Bing Xiao, Xinbo Gao, Dacheng Tao, et al. Photo-sketch synthesis and recognition based on subspace learning. Neurocomputing., Jan. 2010, vol. 73, no. 4-6, pp. 840-852.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |