Colorization of natural images via L1 optimization

@article{Balinsky2009ColorizationON,
  title={Colorization of natural images via L1 optimization},
  author={Alexander A. Balinsky and Nassir Mohammad},
  journal={2009 Workshop on Applications of Computer Vision (WACV)},
  year={2009},
  pages={1-6}
}
  • A. Balinsky, Nassir Mohammad
  • Published 18 May 2009
  • Computer Science, Mathematics
  • 2009 Workshop on Applications of Computer Vision (WACV)
Natural images in the colour space Y UV have been observed to have a non-Gaussian, heavy tailed distribution (called ‘sparse’) when the filter is applied to the chromacity channel U (and equivalently to V ), where w is a weighting function constructed from the intensity component Y [1]. In this paper we develop Bayesian analysis of the colorization problem using the filter response as a regularization term to arrive at a non-convex optimization problem. This problem is convexified using L1… 
Non-linear filter response distributions of natural images and applications in image processing
TLDR
This paper shows that the non-linear filter response distributions of natural colour images are also sparse, and incorporates the statistics of natural images into Bayesian models to illustrate image processing applications in colorization of gray images, compression of colour images and reconstruction of chroma channels from corrupted data.
Sparse natural image statistics and their applications to colorization and compression
Statistical analysis of natural luminance and colour images have shown to display non-Gaussian distributions. We review a recent result on the non-linear sparse filter response distribution observed
Sparse modelling of natural images and compressive sensing
TLDR
A Bayesian treatment to the problem of colorizing natural gray images, and image compression schemes using elements of compressive sensing and sparsity are given and a denoising algorithm that utilises the sparse filter responses as a regular- isation function for the effective attenuation of Gaussian and impulse noise in images is proposed.
Automatic Example-Based Image Colorization Using Location-Aware Cross-Scale Matching
TLDR
This paper proposes a novel cross-scale texture matching method to improve the robustness and quality of the colorization results and considers a class of semantic violation where the statistics of up-down relationships learned from the reference image are violated.
Colorizing Gray Level Images by using Wavelet Filters
TLDR
A new algorithm to produce colored version of gray scale natural still images using artificial neural network to predict RGB channels using the Discrete Wavelet Transform and it is compared to similar algorithm based on discrete cosine transform.
Multi-sparse dictionary colorization algorithm based on the feature classification and detail enhancement
TLDR
A multi-sparse dictionary colorization algorithm based on the feature classification and detail enhancement (CEMDC) is proposed based on this framework, which can achieve a natural colorized effect for a gray-scale image, and it is consistent with the human vision.
Colorization of gray scale natural still images by using ANN to predict the low frequency DCT components of the RGB channels
TLDR
A new algorithm for colorizing gray scale natural still images using artificial neural network (ANN) to predict the low frequency discrete cosine transform (DCT) components of the RGB channels.
Colourization of Low-Light-Level Images Based on Rule Mining
TLDR
The colourization of grey image has been the hotspot research on night-vision technology for a long time and is significant military and civil fields and the human eyes have a high resolution and sensitivity to colourful images.
On semi-supervised learning and sparsity
  • A. Balinsky, Helen Balinsky
  • Mathematics, Computer Science
    2009 IEEE International Conference on Systems, Man and Cybernetics
  • 2009
TLDR
A connection between semi-supervised learning and compressive sampling is established and it is shown that sparsity and compressibility of the learning function can be obtained from heavy-tailed distributions of filter responses or coefficients in spectral decompositions.

References

SHOWING 1-10 OF 17 REFERENCES
Geometry and Color in Natural Images
TLDR
An experimental procedure is proposed and a mathematical proof that the algorithmic procedure is intrinsic, i.e. does not depend asymptotically upon the quantization mesh used for the topographic map, and it is proved its contrast invariance.
Fast image and video colorization using chrominance blending
TLDR
Based on the concepts of luminance-weighted chrominance blending and fast intrinsic distance computations, high-quality colorization results for still images and video are obtained at a fraction of the complexity and computational cost of previously reported techniques.
Colorization using optimization
TLDR
This paper presents a simple colorization method that requires neither precise image segmentation, nor accurate region tracking, and demonstrates that high quality colorizations of stills and movie clips may be obtained from a relatively modest amount of user input.
Non-linear Filter Response Distributions of Natural Colour Images
TLDR
A Generalised Gaussian Distribution (GGD) is fitted to the histogram of the filter response, and the shape parameter (?) to lie within the range 0 1.
Natural Image Colorization
TLDR
An interactive system for users to easily colorize the natural images of complex scenes and employs a smoothness map to guide the incorporation of intensity-continuity and texture-similarity constraints in the design of the labeling algorithm.
Learning to compress images and videos
TLDR
This work presents an intuitive scheme for lossy color-image compression, using the color information from a few representative pixels to learn a model which predicts color on the rest of the pixels, and a simple active learning strategy to choose the representative pixels.
Colorization by example
TLDR
A new method for colorizing grayscale images by transferring color from a segmented example image is presented, rather than relying on a series of independent pixel-level decisions, that attempts to account for the higher-level context of each pixel.
Image compression using sparse colour sampling combined with nonlinear image processing
TLDR
Two recent non-linear, image-processing algorithms are applied to colour image compression and show interesting behaviour, with the second showing a visible improvement in visual quality, over JPEG, at the same compression rate.
Color Image Coding by Colorization Approach
TLDR
A new color image coding scheme called "colorization image coding" is proposed, based on the colorization technique which can colorize a monochrome image by giving a small number of color pixels.
Solving large-scale linear programs by interior-point methods under the Matlab ∗ Environment †
TLDR
This paper describes the implementation of a primal-dual infeasible-interior-point algorithm for large-scale linear programming under the MATLAB environment, and discusses in detail a technique for overcoming numerical instability in Cholesky factorization at the end-stage of iterations in interior-point algorithms.
...
1
2
...