Color Beaver: Bounding Illumination Estimations for Higher Accuracy

Koščević, Karlo; Banić, Nikola; Lončarić, Sven

doi:10.5220/0007393701830190

Cited by 11 publications

(15 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A good starting point to explain why this is so is to mention the experiment conducted in [38] where several Canon camera models were used to take images influenced by several hundreds illuminations. It has been shown how Canon cameras simply restrict most illumination colors to a polygon that encloses the most commonly observed illuminations.…”

Section: General Applicabilitymentioning

confidence: 99%

Unsupervised Learning for Color Constancy

Banić

Lončarić

2018

Proceedings of the 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

Self Cite

View full text Add to dashboard Cite

Most digital camera pipelines use color constancy methods to reduce the influence of illumination and camera sensor on the colors of scene objects. The highest accuracy of color correction is obtained with learning-based color constancy methods, but they require a significant amount of calibrated training images with known ground-truth illumination. Such calibration is time consuming, preferably done for each sensor individually, and therefore a major bottleneck in acquiring high color constancy accuracy. Statistics-based methods do not require calibrated training images, but they are less accurate. In this paper an unsupervised learning-based method is proposed that learns its parameter values after approximating the unknown ground-truth illumination of the training images, thus avoiding calibration. In terms of accuracy the proposed method outperforms all statisticsbased and many learning-based methods. An extension of the method is also proposed, which learns the needed parameters from non-calibrated images taken with one sensor and which can then be successfully applied to images taken with another sensor. This effectively enables inter-camera unsupervised learning for color constancy. Additionally, a new high quality color constancy benchmark dataset with 1707 calibrated images is created, used for testing, and made publicly available. The results are presented and discussed. The source code and the dataset are available at

show abstract

Section: General Applicabilitymentioning

confidence: 99%

Unsupervised Learning for Color Constancy

Banić

Lončarić

2018

Proceedings of the 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

Self Cite

View full text Add to dashboard Cite

show abstract

“…The problem of dependency of illumination estimation methods on the camera sensor was tackled in [24], where two convolutional networks were used for sensor space mapping and illumination estimation, respectively. Other learning-based methods use Bayesian learning [25], color moments [26], gamut mapping [27]- [29], spatial localizations [30], [31], visual information of high level [32], illumination space restrictions [4], [33]- [35], gray pixel detection [36], regression trees with simple color features [37], and others.…”

Section: A Illumination Estimationmentioning

confidence: 99%

“…Such illuminations can corrupt object colors, and if their estimates are imprecise high errors in corrected images can be expected. In [4], it was shown that camera manufacturers bound illuminations to a narrow region in chromaticity space so that chromatic adaptation is never performed with highly colored illuminations. It can be speculated that the cause for this is the inadequacy of the chromatic adaptation model that is unfit for the highly colored illuminations.…”

Section: Introductionmentioning

confidence: 99%

Iterative Convolutional Neural Network-Based Illumination Estimation

2021

View full text Add to dashboard Cite

In the image processing pipelines of digital cameras, one of the first steps is to achieve invariance in terms of scene illumination, namely computational color constancy. Usually, this is done in two successive steps which are illumination estimation and chromatic adaptation. The illumination estimation aims at estimating a three-dimensional vector from image pixels. This vector represents the scene illumination, and it is used in the chromatic adaptation step, which aims at eliminating the bias in image colors caused by the color of the illumination. An accurate illumination estimation is crucial for successful computational color constancy. However, this is an ill-posed problem, and many methods try to comprehend it with different assumptions. In this paper, an iterative method for estimating the scene illumination color is proposed. The method calculates the illumination vector by a series of intermediate illumination estimations and chromatic adaptations of an input image using a convolutional neural network. The network has been trained to iteratively compute intermediate incremental illumination estimates from the original image. Incremental illumination estimates are combined by per element multiplication to obtain the final illumination estimation. The approach is aimed to reduce large estimation errors usually occurring with highly saturated light sources. Experimental results show that the proposed method outperforms the vast majority of illumination estimation methods in terms of median angular error. Moreover, in terms of worst-performing samples, i.e., the samples for which a method errs the most, the proposed method outperforms all other methods by a margin of more than 18% with respect to the mean of estimation errors in the third quartile.

show abstract

“…On the other hand the second group consists of learning-based methods such as gamut mapping (pixel, edge, and intersection based) [12], using high-level visual information [13], natural image statistics [14], Bayesian learning [15], spatio-spectral learning (maximum likelihood estimate, and with gen. prior) [16], simplifying the illumination solution space [17], [18], [19], using color/edge moments [20], using regression trees with simple features from color distribution statistics [21], performing various kinds of spatial localizations [22], [23], using convolutional neural networks [24], [25], [26], [27] and genetic algorithms [28], modelling colour constancy by using the overlapping asymmetric Gaussian kernels with surround pixel contrast based sizes [29], finding paths for the longest dichromatic line produces by specular pixels [30], detecting gray pixels with specific illuminant-invariant measures in logarithmic space [31], channel-wise pooling the responses of double-opponency cells in LMS color space [32], and numerous other. Low-level statistics-based method rely on simple image statistics and therefore, they are fast, computationally cheap, and suitable for hardware implementation.…”

Section: Related Workmentioning

confidence: 99%

“…Third, while the illumination color can be arbitrary when artificial light sources are used, in practice the digital cameras mostly focus only on a restricted set of illuminations, usually the ones whose chromaticities are close to the ones of the black body radiation colors [28]. Therefore, when looking for an answer to the first two questions, it would be useful to give it individually for the case when a large variety of illuminations, e.g.…”

Section: Proposed Experimentsmentioning

confidence: 99%

On Some Desired Properties of Data Augmentation by Illumination Simulation for Color Constancy

Banić¹,

Koščević²,

Subašić³

et al. 2020

Computer Science &Amp; Information Technology (CS &Amp; IT)

Self Cite

View full text Add to dashboard Cite

Computational color constancy is used in almost all digital cameras to reduce the influence of scene illumination on object colors. Many of the highly accurate published illumination estimation methods use deep learning, which relies on large amounts of images with known ground-truth illuminations. Since the size of the appropriate publicly available training datasets is relatively small, data augmentation is often used also by simulating the appearance of a given image under another illumination. Still, there are practically no reports on any desired properties of such simulated images or on the limits of their usability. In this paper, several experiments for determining some of these properties are proposed and conducted by comparing the behavior of the simplest illumination estimation methods on images of the same scenes obtained under real illuminations and images obtained through data augmentation. The experimental results are presented and discussed.

show abstract

Color Beaver: Bounding Illumination Estimations for Higher Accuracy

Cited by 11 publications

References 0 publications

Unsupervised Learning for Color Constancy

Unsupervised Learning for Color Constancy

Iterative Convolutional Neural Network-Based Illumination Estimation

On Some Desired Properties of Data Augmentation by Illumination Simulation for Color Constancy

Contact Info

Product

Resources

About