A locally adaptive L1−L2 norm for multi-frame super-resolution of images with mixed noise and outliers
Introduction
Due to the limitation of solid-state sensors such as CCD or CMOS, the digital images acquired may have a limited spatial resolution, thus reduce their practical value. In addition, digital images may be degraded by blurring or noise because the information each pixel records is easily polluted during the acquisition and transmission procedure. Super-resolution is a technique which takes resolution limitation and regular image degradation into consideration at the same time. Using the redundant information between multi-frame images with a relative sub-pixel motion, the super-resolution technique can construct a higher-resolution image or sequence, and it has been widely applied with medical images [1], remote sensing images [2], [3], [4], and video surveillance [5], [6].
The multi-frame super-resolution technique has been developed for almost three decades [7]. The frequency domain approaches were first addressed; however, these methods [7], [8], [9] are extremely sensitive to model errors, and have difficulty in including prior knowledge to solve the model. Therefore, methods in the spatial domain have become more popular in recent years [3], [10], [11], [12], [13]. These methods include iterative back projection (IBP) [10], projection onto convex sets (POCS) [11], and a group of Bayesian-based probability and statistical methods [3], [12], [13]. Among the methods, the Bayesian-based methods have the advantages of adding a prior and simultaneously handling the high-resolution (HR) image estimation. Furthermore, the model parameters such as the motion vector and blur kernel are estimated iteratively [12]. As a result, the Bayesian-based methods have become the most widely used. These methods in the spatial domain regard super-resolution as an ill-posed inverse problem. Supposing the degradation procedure during image acquisition involves warping, blurring, downsampling, and noise (Fig. 1), then the universal observation model is usually described as follows [14], [15]:where is the observed image of size , and is the desired HR image of size , which is determined by the downsampling factor. , , and are, respectively, the downsampling, blurring, and motion operators, and is the additive noise. If and are excluded, it is a model for image restoration dealing with problems with only noise and blurring.
The standard regularized minimizing function usually consists of a data fidelity term describing the model error, and a regularization to constrain the model to achieve a robust solution, which is often described as [12], [16], [17]
In (2), K represents the number of low-resolution (LR) images. is the regularization, which is generally chosen from Tikhonov [18], Huber–Markov random field (HMRF) [19], or total variation (TV) family [12], [15], [20], and λ is the regularized parameter which controls the tradeoff between the two terms. For the data fidelity term, the norm is widely used, but it only performs well for model errors satisfying a traditional Gaussian distribution [21]. It has been proved that the norm is more appropriate for impulse noise and motion outliers in image inverse problems. The reason for the robustness of when handling impulse errors is based on the fact that the model results in a pixel-wise mean, while the model results in a pixel-wise median of all the measurements after motion compensation [15], [16], [22].
In the image processing field, Gaussian-type noise is the most commonly assumed because the noise generated in image acquisition usually satisfies a Gaussian distribution. However, in many applications, practical systems suffer from impulse noise/outliers, which are typically caused by malfunctioning arrays in camera sensors, faulty memory locations in hardware, or transmission in a noisy channel [23], [24]. In the image super-resolution problem, motion outliers should be specially considered because some pixels in one frame may be unobservable in the other frames [25], [26], [27]. Considering the complicated degradation of images, as with mixed noise and motion outliers, both the and models have their own advantages and disadvantages. Therefore, some researchers have tried to combine the advantages of the and norms for the fidelity [28], [29], [30]. However, the combination of and may result in new challenges. The hybrid-norm problem is not a standard linear problem to solve, so an efficient optimization method should be considered. In addition, the reconstruction capability of the regularized model hinges on the selection of , and different norm constraints should correspond to different weights for the tradeoff between the fidelity and the regularization. The traditional methods usually tune the weight for the hybrid norm manually, which is time consuming [31]; thus, adaptive weight estimation is an issue worth considering.
In order to deal with super-resolution with mixed noise and/or motion outliers, we propose a variational model employing a regularized framework with a locally adaptive fidelity norm. Specifically, different norm values in the data fidelity for different pixel locations are determined according to the detection results for the impulse noise and motion outliers. The norm is employed for pixels with impulse noise and motion outliers, and the norm is used for the other pixels. To balance the difference in the constraint strength between the norm and the norm, a strategy to adaptively estimate the weighted parameter is put forward.
The remainder of this paper is organized as followed. We introduce the adaptive model construction and optimization in Section 2. The norm selection is described in Section 2.1, in which detection for impulse noise and motion outliers is presented separately. The solving strategy and the method of adaptively determining the weight for the norm-adaptive data fidelity term are given in Section 2.2. We present the experiments for super-resolution under mixed noise conditions, including real video sequence images with moving objects, in Section 3, along with a discussion of the comparative results and parameters. Section 4 is the conclusion.
Section snippets
The locally adaptive super-resolution method
As mentioned before, the model in (2) with is usually robust for the Gaussian noise case, while is suitable for impulse error and/or outliers. Due to the complicated imaging process, there may be mixed noise (mainly Gaussian plus impulse noise) [29], [31], [32], [33], [34] and moving objects, which may lead to outliers in the image sequences [25], [26], [27]. To address these problems, we employ an adaptive norm framework. Driven by the detection map of the impulse noise and
Experiments
The experiments consist of two parts, to test the effectiveness of the proposed model in handling super-resolution for images with mixed noise and/or motion outliers. In the first part (Section 3.1), three commonly used images with different resources, and contaminated by different levels of mixed noise, are chosen as the test images. The experiments in this part are mainly designed to test the effectiveness of the model in suppressing mixed noise during super-resolution. The experiments in the
Conclusion
Super-resolution reconstruction becomes more complicated when there is mixed corruption, such as mixed noise and/or multiple independent moving objects. To take the mixed degradation factor into account and solve it with a generalized model, we propose a locally adaptive norm super-resolution method to make use of the different norms’ advantages for different types of model error. The proposed method adaptively selects a pixel-based norm, based on the detection results of the impulse
Acknowledgments
The authors would like to thank the editors and the anonymous reviewers for their valuable suggestions. This research is supported by Major State Basic Research Development Program (2011CB707105), National Natural Science Foundation of China (41271376), Program for Changjiang Scholars and Innovative Research Team in University (IRT1278) and Ph.D. Programs Foundation of Ministry of Education of China under Grant 20130141120001.
References (58)
- et al.
Infrared surveillance image super resolution via group sparse representation
Opt. Commun.
(2013) - et al.
A super-resolution reconstruction algorithm for hyperspectral images
Signal Process.
(2012) - et al.
Classification-based video super-resolution using artificial neural networks
Signal Process.
(2013) - et al.
Improving resolution by image registration, CVGIP
Graph. Models Image Process.
(1991) - et al.
Super-resolution image reconstruction techniques: trade-offs between the data-fidelity and regularization terms
Inf. Fusion
(2012) - et al.
Local object-based super-resolution mosaicing from low-resolution video
Signal Process.
(2011) - et al.
Motion based video super resolution using edge directed interpolation and complex wavelet transform
Signal Process.
(2013) - et al.
Restoration of images corrupted by mixed Gaussian-impulse noise via l1-l0 minimization
Pattern Recognit.
(2011) - et al.
A robust multiframe super-resolution algorithm based on half-quadratic estimation with modified BTV regularization
Digital Signal Process.
(2013) - et al.
A restoration algorithm for images contaminated by mixed Gaussian plus random-valued impulse noise
J. Vis. Commun. Image Represent.
(2013)
Restoration of images corrupted by Gaussian and uniform impulsive noise
Pattern Recognit.
Subpixel motion estimation for super-resolution image sequence enhancement
J. Vis. Commun. Image Represent.
Nonlinear structure tensors
Image Vis. Comput.
A multi-frame image super-resolution method
Signal Process.
Fractional order total variation regularization for image super-resolution
Signal Process.
Markov random field models for non-quadratic regularization of complex SAR images
IEEE J. Sel. Top. Appl. Earth Obser. Remote Sens.
A Bayesian restoration approach for hyperspectral images
IEEE Trans. Geosci. Remote Sens.
A MAP approach for joint motion estimation, segmentation, and super resolution
IEEE Trans. Image Process.
Multi-frame image restoration and registration
Adv. Comput. Vis. Image Process.
Recursive reconstruction of high resolution image from noisy undersampled multiframes
IEEE Trans. Acoust. Speech Signal Process.
A total variation regularization based super-resolution reconstruction algorithm for digital video
EURASIP J. Adv. Signal Process.
Super-resolution image reconstruction: a technical overview
IEEE Signal Process. Mag.
Fast and robust multiframe super resolution
IEEE Trans. Image Process.
Application of Tikhonov regularization to super-resolution reconstruction of brain MRI images
Med. Imaging Inf.
Cited by (56)
Image identification of cashmere and wool fibers based on the improved Xception network
2022, Journal of King Saud University - Computer and Information SciencesMLSR: Missing information-based fidelity and learned regularization for single-image super-resolution
2022, Computers and Electrical EngineeringReal-world single image super-resolution: A brief review
2022, Information FusionCitation Excerpt :On the whole, VSR aims to improve the spatial resolution (known as spatial VSR) [13–21] or the frame rate (known as temporal VSR) [22–29] of the observed video. ISR can be further classified into multi-frame image super-resolution (MISR) [30–38] and single image super-resolution (SISR) [39–82]. MISR refers to reconstructing an HR image via fusing the complementary information in a series of correlated images of the same scene [30–38], while SISR generates an HR image from one LR observation [39–82].
A multi-frame super-resolution based on new variational data fidelity term
2020, Applied Mathematical ModellingMulti-frame super-resolution reconstruction based on mixed Poisson–Gaussian noise
2020, Signal Processing: Image Communication