SLIC super-pixels for multi-resolution compressive spectral imaging reconstruction

Uso de super pixeles SLIC para la reconstrucción de imágenes espectrales adquiridas mediante muestreo compresivo

By: H. García, C. V. Correa, H. Argüello

Main Information

Volume:
Vol.51-N3 / 2018 - Ordinario
Section:
Spectroscopy
Pages:
50304:1-10
DOI:
http://doi.org/10.7149/OPA.51.3.50304
Type:
Research papers / Trabajos de investigación
Language:
English
Attachments:
Keywords:
Multi-resolution, super-pixels, single pixel camera, compressive spectral imaging.

Multi-resolución múltiple, super pixeles, cámara de un único píxel, muestreo compresivo, imágenes espectrales.
PDF file download

Abstract

Spectral imaging (SI) is widely used in different applications involving material identification since it contains both spatial (x,y) and spectral information (λ). However, traditional SI scanning methods involve massive amounts of data, which increase the cost of storing and processing. Compressive sensing (CS) theory has been applied in SI, such that the underlying data cube can be recovered from a reduced number of measures. Reconstructions are obtained by l2-l1 norm-based algorithms whose computational complexity grows in proportion to the number of unknowns. In this paper, a multi-resolution reconstruction model based on the simple linear iterative clustering (SLIC) is proposed to reduce the number of unknown values to recover. Simulation results show that the proposed method is up to 86% faster than the full-resolution reconstructions. Additionally, MR approach obtains more accurate reconstructions with improvements of up to 12dB of PSNR.


Las imágenes espectrales (SI) se utilizan comúnmente en diferentes aplicaciones que involucran identificación de materiales debido a que contienen información espacial (x,y) y espectral (λ). Sin embargo, los métodos de adquisición tradicionales emplean gran cantidad de datos, lo que conlleva a altos costos de almacenamiento y procesamiento. Muestreo compresivo (CS) se ha aplicado en SI con el fin de recuperar el cubo de datos a partir de un número reducido de medidas. Las reconstrucciones se obtienen mediante algoritmos basados en la norma l2-l1 cuya complejidad computacional crece en proporción al número de incógnitas. En este trabajo, se propone un modelo de reconstrucción de resolución múltiple basado en el algoritmo iterativo de agrupamiento lineal simple (SLIC) para disminuir el número de incógnitas a recuperar. Los resultados de simulación muestran que el método propuesto es hasta un 86% más rápido que las reconstrucciones de resolución completa. Además, el enfoque de MR obtiene una reconstrucción más precisa en hasta 12dB de PSNR.

References

[1] P. Ghamisi, N. Yokoya, J. Li, W. Liao, S. Liu, J. Plaza, B. Rasti, and A. Plaza, "Advances in Hyperspectral Image and Signal Processing: A Comprehensive Overview of the State of the Art," IEEE Geosci. Remote Sens. Mag., 5, pp. 37-78, (2017).

[2] D. L. Donoho, "Compressed sensing," IEEE Trans. Inf. Theory, 52, , pp. 1289-1306, (2006).

[3] X. Lin, Y. Liu, J. Wu, and Q. Dai, "Spatial-spectral encoded compressive hyperspectral imaging," ACM Trans. Graph., 33, , p. 233:1-233:11, (2014).

[4] A. Wagadarikar, R. John, R. Willett, and D. Brady, "Single disperser design for coded aperture snapshot spectral imaging," Appl. Opt., 47, pp. B44-B51, (2008).

[5] X. Lin, G. Wetzstein, Y. Liu, and Q. Dai, "Dual-Coded Compressive Hyper-Spectral Imaging," Class. Opt. 2014, OSA Tech. Dig. (Optical Soc. Am. 2014), 39, pp. 2044-2047, (2014).

[6] X. Cao, H. Du, X. Tong, Q. Dai, and S. Lin, "A prism-mask system for multispectral video acquisition," IEEE Trans. Pattern Anal. Mach. Intell., 33, pp. 2423-2435, (2011).

[7] M. Dadkhah, J. M. Deen, and S. Shirani, "Compressive sensing image sensors-hardware implementation," Sensors (Switzerland), 13, pp. 4961-4978, (2013).

[8] M. F. Duarte, M. a. Davenport, D. Takhar, J. N. Laska, T. S. T. Sun, K. F. Kelly, and R. G. Baraniuk, "Single-Pixel Imaging via Compressive Sampling," IEEE Signal Process. Mag., 25, pp. 1-19, (2008).

[9] M. A. T. Figueiredo, R. D. Nowak, and S. J. Wright, "Gradient Projections For Sparse Reconstruction: Application to Compressed Sensing and Other Inverse Problems," J. Sel. Top. Signal Process. IEEE, 1, , pp. 586-598, (2007).

[10] T. T. Cai and L. Wang, "Orthogonal Matching Pursuit for Sparse Signal Recovery With Noise," IEEE Trans. Inf. theory, 57, pp. 4680-4688, (2011).

[11] J. M. Bioucas-dias and M. A. T. Figueiredo, "A New TwIST : Two-Step Iterative Shrinkage / Thresholding Algorithms for Image Restoration," IEEE Trans. Image Process., 16, pp. 2992-3004, (2007).

[12] Itu-t, "ITU-T Rec. H.265 (10/2014) High efficiency video coding," 265, (2014).

[13] R. Achanta, A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Süsstrunk, "SLIC Superpixels Compared to State-of-the-Art Superpixel Methods," Pattern Anal. Mach. Intell. IEEE Trans., 34, pp. 2274-2282, (2011).

[14] J. Chen, Z. Li, S. Member, and B. Huang, "Linear Spectral Clustering Superpixel," IEEE Trans. Image Process., 26, pp. 3317-3330, (2017).

[15] H. Garcia, C. V Correa, O. Villarreal, S. Pinilla, and H. Arguello, "Multi-Resolution Reconstruction Algorithm for Compressive Single Pixel Spectral Imaging," in 25th European Signal Processing Conference (EUSIPCO), pp. 498-502 (2017).

[16] A. Said and W. A. Pearlman, "An image multiresolution representation for lossless and lossy compression," IEEE Trans. Image Process., 5, pp. 1303-1310, (1996).

[17] A. Gonzalez, H. Jiang, G. Huang, and L. Jacques, "Multi-resolution Compressive Sensing Reconstruction," in 2016 IEEE International Conference on Image Processing (ICIP), pp. 1-5, (2016).

[18] X. Wang and J. Liang, "Multi-Resolution Compressed Sensing via Approximate Message Passing," IEEE Trans. Comput. imaging, 2, pp. 1-31, (2015).

[19] Y. August, C. Vachman, Y. Rivenson, and A. Stern, "Compressive hyperspectral imaging by random separable projections in both the spatial and the spectral domains.," Appl. Opt., 52, pp. D46-54, (2013).

[20] A. C. Sankaranarayanan, L. Xu, C. Studer, Y. Li, K. Kelly, and R. G. Baraniuk, "Video Compressive Sensing for Spatial Multiplexing Cameras using Motion-Flow Models," (2015).

[21] A. Levinshtein, A. Stere, K. N. Kutulakos, D. J. Fleet, S. J. Dickinson, and K. Siddiqi, "TurboPixels: Fast superpixels using geometric flows," IEEE Trans. Pattern Anal. Mach. Intell., 31, 2290-2297, (2009).

[22] A. Vedaldi and S. Soatto, "Quick shift and kernel methods for mode seeking," Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), 5305 LNCS, no. PART 4, pp. 705-718, (2008).

[23] R. Achanta, A. Shaji, K. Smith, A. Lucchi, P. Fua, and S. Susstrunk, "SLIC Superpixels," (2010).

[24] F. Yasuma, T. Mitsunaga, D. Iso, and S. K. Nayar, "Generalized Assorted Pixel Camera : Postcapture Control of Resolution , Dynamic Range , and Spectrum," IEEE Trans. Image Process., 19, pp. 2241-2253, (2010).

[25] C. V. Correa Pugliese, D. F. Galvis Carreño, and H. Arguello Fuentes, "Sparse representations of dynamic scenes for compressive spectral video sensing," Dyna, 83, pp. 42-51, (2016).

[26] M. F. Duarte and R. G. Baraniuk, "Kronecker compressive sensing," IEEE Trans. Image Process., 21, pp. 494-504, (2012).