Computationally Efficient Data and Application Driven Color Transforms for the Compression and Enhancement of Images and Video

An important step in color image or video coding and enhancement is the linear transformation of input (typically red-green-blue (RGB)) data into a color space more suitable for compression, subsequent analysis, or visualization. The choice of this transform becomes even more critical when operating in distributed and low-computational power environments, such as visual sensor networks or remote sensing. Data-driven transforms are rarely used due to increased complexity. Most schemes adopt fixed transforms to decorrelate the color channels which are then processed independently. Here we propose two frameworks to find appropriate data-driven transforms in different settings. The first, named approximate Karhunen–Loeve Transform (aKLT), performs comparable to the KLT at a fraction of the computational complexity, thus favoring adoption on sensors and resource-constrained devices. Furthermore, we consider an application-aware setting in which an expert system (e.g., a classifier) analyzes imaging data at the receiver’s end. In a compression context, distortion may jeopardize the accuracy of the analysis. Since the KLT is not optimal in this setting, we investigate formulations that maximize post-compression expert system performance. Relaxing decorrelation and energy compactness constraints, a second transform can be obtained offline with supervised learning methods. Finally, we propose transforms that accommodate both constraints, and are found using regularized optimization.

[1]  Pengwei Hao,et al.  Reversible integer KLT for progressive-to-lossless compression of multiple component images , 2003, Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429).

[2]  Zhe Liu,et al.  Structured covariance principal component analysis for real-time onsite feature extraction and dimensionality reduction in hyperspectral imaging. , 2014, Applied optics.

[3]  Natasha Gelfand,et al.  A survey of image retargeting techniques , 2010, Optical Engineering + Applications.

[4]  Keinosuke Fukunaga,et al.  Introduction to statistical pattern recognition (2nd ed.) , 1990 .

[5]  Harry Wechsler,et al.  Color image compression using PCA and backpropagation learning , 2000, Pattern Recognit..

[6]  Tomer Hertz,et al.  Learning a Mahalanobis Metric from Equivalence Constraints , 2005, J. Mach. Learn. Res..

[7]  Mircea Andrecut,et al.  Parallel GPU Implementation of Iterative PCA Algorithms , 2008, J. Comput. Biol..

[8]  Gerald J. Agin Computer Vision Systems for Industrial Inspection and Assembly , 1980, Computer.

[9]  C. Rubinstein,et al.  Statistical Dependence Between Components of a Differentially Quantized Color Signal , 1972, IEEE Trans. Commun..

[10]  C. B. Rubinstein,et al.  Digital Coding of Color Video Signals - A Review , 1977, IEEE Transactions on Communications.

[11]  R. E. Walker,et al.  Color enhancement of highly correlated images. I - Decorrelation and HSI contrast stretches. [hue saturation intensity , 1986 .

[12]  Alan C. Bovik,et al.  Mean squared error: Love it or leave it? A new look at Signal Fidelity Measures , 2009, IEEE Signal Processing Magazine.

[13]  B. Penna,et al.  A New Low Complexity KLT for Lossy Hyperspectral Data Compression , 2006, 2006 IEEE International Symposium on Geoscience and Remote Sensing.

[14]  Pengwei Hao,et al.  Matrix factorizations for reversible integer mapping , 2001, IEEE Trans. Signal Process..

[15]  Henryk Palus,et al.  Representations of colour images in different colour spaces , 1998 .

[16]  Sotirios A. Tsaftaris,et al.  Unsupervised and supervised approaches to color space transformation for image coding , 2014, 2014 IEEE International Conference on Image Processing (ICIP).

[17]  Sotirios A. Tsaftaris,et al.  Application-aware image compression for low cost and distributed plant phenotyping , 2013, 2013 18th International Conference on Digital Signal Processing (DSP).

[18]  Yücel Altunbasak,et al.  Chrominance Edge Preserving Grayscale Transformation with Approximate First Principal Component for Color Edge Detection , 2007, 2007 IEEE International Conference on Image Processing.

[19]  T. W. Anderson An Introduction to Multivariate Statistical Analysis , 1959 .

[20]  David Zhang,et al.  Two-stage image denoising by principal component analysis with local pixel grouping , 2010, Pattern Recognit..

[21]  Stephen P. Boyd,et al.  Graph Implementations for Nonsmooth Convex Programs , 2008, Recent Advances in Learning and Control.

[22]  Michal Majchrowicz,et al.  Application of General-Purpose Computing on Graphics Processing Units for Acceleration of Basic Linear Algebra Operations and Principal Components Analysis Method , 2013, ICMMI.

[23]  Nicholas J. Higham,et al.  Functions of matrices - theory and computation , 2008 .

[24]  I. Jolliffe Principal Component Analysis , 2002 .

[25]  Roumen Kountchev,et al.  New method for adaptive karhunen-loeve color transform , 2009, 2009 9th International Conference on Telecommunication in Modern Satellite, Cable, and Broadcasting Services.

[26]  Manohar N. Murthi,et al.  Quantization for classification accuracy in high-rate quantizers , 2011, 2011 Digital Signal Processing and Signal Processing Education Meeting (DSP/SPE).

[27]  Detlev Marpe,et al.  Macroblock-Adaptive Residual Color Space Transforms for 4:4:4 Video Coding , 2006, 2006 International Conference on Image Processing.

[28]  Aggelos K. Katsaggelos,et al.  Low-Complexity Tracking-Aware H.264 Video Compression for Transportation Surveillance , 2011, IEEE Transactions on Circuits and Systems for Video Technology.

[29]  Gary J. Sullivan,et al.  Lifting-based reversible color transformations for image compression , 2008, Optical Engineering + Applications.

[30]  Ying Chen,et al.  Optimal Transform in Perceptually Uniform Color Space and Its Application in Image Coding , 2004, ICIAR.

[31]  Wenming Zheng,et al.  Foley-Sammon optimal discriminant vectors using kernel approach , 2005, IEEE Trans. Neural Networks.

[32]  Yuhua Peng,et al.  A Fast Algorithm for YCbCr to RGB Conversion , 2007, IEEE Transactions on Consumer Electronics.

[33]  Ingeborg Tastl,et al.  Comparison between Different Color Transformations for the JPEG 2000 , 2000, PICS.

[34]  Qian Du,et al.  Low-Complexity Principal Component Analysis for Hyperspectral Image Compression , 2008, Int. J. High Perform. Comput. Appl..

[35]  Inderjit S. Dhillon,et al.  Information-theoretic metric learning , 2006, ICML '07.

[36]  Ashraf A. Kassim,et al.  Embedded color image coding using SPIHT with partially linked spatial orientation trees , 2003, IEEE Trans. Circuits Syst. Video Technol..

[37]  Shohreh Kasaei,et al.  Color PCA eigenimages and their application to compression and watermarking , 2008, Image Vis. Comput..

[38]  Kilian Q. Weinberger,et al.  Distance Metric Learning for Large Margin Nearest Neighbor Classification , 2005, NIPS.

[39]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[40]  Touradj Ebrahimi,et al.  The JPEG 2000 still image compression standard , 2001, IEEE Signal Process. Mag..

[41]  R. Fisher THE USE OF MULTIPLE MEASUREMENTS IN TAXONOMIC PROBLEMS , 1936 .

[42]  P. Schönemann,et al.  A generalized solution of the orthogonal procrustes problem , 1966 .

[43]  Huifang Sun,et al.  Image and Video Compression for Multimedia Engineering: Fundamentals, Algorithms, and Standards , 1999 .

[44]  Fatih Murat Porikli,et al.  Fast Construction of Covariance Matrices for Arbitrary Size Image Windows , 2006, 2006 International Conference on Image Processing.

[45]  Jing-Yu Yang,et al.  Face recognition based on the uncorrelated discriminant transformation , 2001, Pattern Recognit..

[46]  Sotirios A. Tsaftaris,et al.  Image-based plant phenotyping with incremental learning and active contours , 2014, Ecol. Informatics.

[47]  M. Porat,et al.  Does decorrelation really improve color image compression , 2005 .

[48]  Alan R. Gillespie,et al.  Color enhancement of highly correlated images. II. Channel ratio and “chromaticity” transformation techniques , 1987 .

[49]  Markku Hauta-Kasari,et al.  Real-time PCA calculation for spectral imaging (using SIMD and GP-GPU) , 2010, Journal of Real-Time Image Processing.

[50]  John W. Sammon,et al.  An Optimal Set of Discriminant Vectors , 1975, IEEE Transactions on Computers.

[51]  Pengwei Hao,et al.  Comparative study of color transforms for image coding and derivation of integer reversible color transform , 2000, Proceedings 15th International Conference on Pattern Recognition. ICPR-2000.

[52]  R.K. Kouassi,et al.  Application of the Karhunen-Loeve transform for natural color images analysis , 1997, Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136).

[53]  Andy C. Downton,et al.  Karhunen-Loève Transform: An Exercise in Simple Image-Processing Parallel Pipelines , 1997, Comput. Artif. Intell..

[54]  Charles A. Poynton A Guided Tour of Colour Space , 1995 .

[55]  Max A. Viergever,et al.  Ridge-based vessel segmentation in color images of the retina , 2004, IEEE Transactions on Medical Imaging.

[56]  Fatih Celiker,et al.  Fast Color Space Transformations Using Minimax Approximations , 2010, ArXiv.

[57]  Woo-Shik Kim,et al.  A new color transform for RGB coding , 2004, 2004 International Conference on Image Processing, 2004. ICIP '04..

[58]  Berthold K. P. Horn,et al.  Closed-form solution of absolute orientation using orthonormal matrices , 1988 .

[59]  W. Pratt Spatial Transform Coding of Color Images , 1971 .

[60]  Brian Kulis,et al.  Metric Learning: A Survey , 2013, Found. Trends Mach. Learn..

[61]  Ran Ginosar,et al.  Spatio-chromatic model for colour image processing , 1994, Proceedings of 12th International Conference on Pattern Recognition.

[62]  Hanno Scharr,et al.  Annotated Image Datasets of Rosette Plants , 2014 .

[63]  Qian Du,et al.  Hyperspectral Image Compression Using JPEG2000 and Principal Component Analysis , 2007, IEEE Geoscience and Remote Sensing Letters.