Academia.eduAcademia.edu

Outline

Title
Abstract
FAQs
Figures
Introduction
Conclusion
References

Image Super Resolution-A Survey

Profile image of Amisha ShahAmisha Shah

Abstract

Improving image quality has always been an issue of image technology. Enhancing the quality of image is a continuous ongoing process. For some applications it becomes essential to have best quality of image such as in forensic department, where in order to retrieve maximum possible information image has to be enlarged in terms of size. For example sometimes in forensic investigations either criminal face or in video surveillance a licences number plate, increased image size helps to extract minute information embedded in the image. During this exercise beyond a certain limit, the enlarged image results in a blurred picture. Most possible cause of this problem is the hardware limitation, primarily which includes the sampling rate of a Charge- Coupled Device [CCD]. Especially this becomes very critical while capturing a high speed moving object. In such case pre/post-image processing is required which in turn restores the High resolution (HR) / Super resolution (SR) image(s) from the captured low resolution (LR) image(s). Such obtained high quality images have also a concern in satellite imaging, medical science, high Definition Television (HD TV), etc. There are various techniques to attain an image with higher resolution. In this paper some of the approaches of super resolution are discussed.

FAQs

sparkles

AI

What explains the efficiency of frequency domain approaches in super resolution?add

Frequency domain methods leverage the shifting property of the Fourier transform to enhance resolution, demonstrating improved results by compensating for motion blur through interpolation techniques.

How do learning-based approaches enhance computational performance in image super resolution?add

The learning-based methods partition high-frequency patches into classes, using specialized algorithms for edge areas and interpolation in flat regions, significantly reducing computational complexity.

What limitations arise with the Maximum a Posteriori estimation in super resolution techniques?add

While MAP provides robustness against noise, it is highly sensitive to contamination and requires prior knowledge which may not always be available, impacting overall performance.

When did iterative back projection approaches first gain traction in image super resolution?add

The iterative back projection method, initially proposed by Irani-Peleg and Komatsu et al., became notable for enhancing resolution using multiple low-resolution images in the early 2000s.

Why does the projection onto convex sets method have increased computational costs?add

The POCS method requires prior information and iterative projections to achieve a valid image solution, leading to higher computational overhead compared to simpler methods.

Figures (6)
Figurel. (a) Resultant HR image captured from four different cameras (b) Multiple frames captured by a video camera. restoration is used to recover a degraded image without changing its dimension.
Figurel. (a) Resultant HR image captured from four different cameras (b) Multiple frames captured by a video camera. restoration is used to recover a degraded image without changing its dimension.
The basic premise for increasing the spatial resolution in SR techniques is the availability of multiple LR images captured from the same scene [1]. These LR images must be aliased and have sub pixel shift seeing that the LR images having shift of integer units does not enclose new information which can be used for SR reconstruction as revealed in fig.3. The SR image reconstruction consists of three stages as shown in the fig.4. By means of image registration, all pixels from available LR images can be mapped back onto the reference image. Then via interpolation a uniformly spaced up sampled image is obtained. In the end, image restoration is applied to the up-sampled image to remove the effect of sensor PSF blur and noise [4].
The basic premise for increasing the spatial resolution in SR techniques is the availability of multiple LR images captured from the same scene [1]. These LR images must be aliased and have sub pixel shift seeing that the LR images having shift of integer units does not enclose new information which can be used for SR reconstruction as revealed in fig.3. The SR image reconstruction consists of three stages as shown in the fig.4. By means of image registration, all pixels from available LR images can be mapped back onto the reference image. Then via interpolation a uniformly spaced up sampled image is obtained. In the end, image restoration is applied to the up-sampled image to remove the effect of sensor PSF blur and noise [4].
Figure 4 Block diagram of constructing a SR image. Super resolution techniques are broadly classified as techniques based on frequency domain or techniques based on
Figure 4 Block diagram of constructing a SR image. Super resolution techniques are broadly classified as techniques based on frequency domain or techniques based on
Various regularization parameters In learning based approach, the relationship between an LR image and its corresponding HR image is examined via a pair of LR and HR patches. The training data then is used to predict the higher-resolution image. In order to apply the super resolution approach in real time, the super resolving time should be reduce. Xie Qinlan et al. proposed an improved example based single-image super-resolution method in [8] .To enhance the computing performance, the author had proposed a classifying method to divide the high-frequency patches of low-resolution image into different classes. For the edge-areas of the low-resolution, the routine example-based image super resolution algorithm can be used to implement the local and
Various regularization parameters In learning based approach, the relationship between an LR image and its corresponding HR image is examined via a pair of LR and HR patches. The training data then is used to predict the higher-resolution image. In order to apply the super resolution approach in real time, the super resolving time should be reduce. Xie Qinlan et al. proposed an improved example based single-image super-resolution method in [8] .To enhance the computing performance, the author had proposed a classifying method to divide the high-frequency patches of low-resolution image into different classes. For the edge-areas of the low-resolution, the routine example-based image super resolution algorithm can be used to implement the local and
Figure 5 A pictorial example of iterative back projection method. (g“ —g) in a HR image and c is a normalizing constant 1171.
Figure 5 A pictorial example of iterative back projection method. (g“ —g) in a HR image and c is a normalizing constant 1171.

Related papers

A Survey on Image Super- Resolution Techniques for Image Reconstruction
Josephin Nivetha

2016

Image reconstruction techniques are used to create a two dimensional and three dimensional images. For image reconstruction different methods are used such as back projection filters. Image reconstruction is commonly referd to as restoration of missing parts. Super-resolution technique aims to increase the resolution of the limits of the original image or video. It is used to extract the lost details of an image when it was up scaled. Interpolation based SR, Example based SR and Multi image based SR are the main techniques for reconstructing a super resolution image from LR image. Resolution refers to denote the number of pixels in an image and it is measured in pixel Per Inch (PPI). SR technique reduces the image’s blurring and used in many image processing applications. SR technique applied in an improvement of test images, compressed video and image enhancement, medical imaging process and satellite and aerial imaging. SR is a badly postured issue on the grounds that every LR pix...

View PDFchevron_right
Super-Resolution Image Reconstruction: A Technical Overview
Yana Alshevskaya
View PDFchevron_right
IMPROVED SUPER-RESOLUTION METHOD AND ITS ACCELERATION
Vaclav Skala

A brief introduction to spatial-domain Super-Resolution methods, i.e. spatial resolution enhancement methods that create one high-resolution image from a series of low-resolution images shifted by a sub-pixel distance, is given. An improvement applicable to some of existing Super-Resolution methods is presented. Principles of digital photography processing techniques are exploited in order to reduce error in the Super-Resolution process. Enhanced registration method applicable to full color images is proposed and results of its hardware implementation are presented.

View PDFchevron_right
Intensification of Resolution in the Realm of Digital Imaging
Shivangi Jindal, IJCERT Publications

High resolution (HR) images have great importance in many areas, such as astronomy, medical, video surveillance, etc. These HR images give some additional details which are of great significance for analysis in various applications. This paper investigates mainly on various methods of super resolution that are existing, and putting it all together for a literature survey. Scope of this study mainly focuses on the different available techniques of image processing to get high resolution images.

View PDFchevron_right
ANALYSIS OF SINGLE FRAME SUPER RESOLUTION METHODS
Editor International Journal of Engineering Development and Research IJEDR

INTERNATIONAL JOURNAL OF ENGINEERING DEVELOPMENT AND RESEARCH (IJEDR) (ISSN:2321-9939), 2014

The Image Quality can be most often measured in terms of Resolution. The clarity of the Image can be determined by Resolution which means higher the resolution, the image can be more clearer which is most often required in most of the applications. It can be achieved by use of Good Sensors ad optics, but it can be very expensive and also limit way of pixel density within Image. Instead of that we can use image processing methods to obtain High resolution image from low resolution image which can very effective and cheap solution. This Kind of Image Enhancement is called Super Resolution Image Reconstruction. This paper focuses on the definition, implementation and analysis on well-known techniques of super resolution. The comparison and analysis are the main concerns to understand the improvements of the super resolution methods over single frame interpolation techniques. In addition, the comparison also gives us an insight to the practical uses of super resolution methods. As a result of the analysis, the critical examination of the techniques and their performance evaluation are achieved. Super Resolution is particularly useful in forensic imaging, where the extraction of minute details in an image can help to solve a major crime cases. Super-resolution image restoration has been one of the most important research areas in recent years which goals to obtain a high resolution (HR) image from low resolutions (LR) blurred, noisy, under sampled and displaced image.

View PDFchevron_right
A SURVEY ON SUPER RESOLUTION TECHNIQUES
IJESRT Journal

The main aim of super resolution image is to reconstruct a high-resolution(HR) image from low resolution images. Nowadays super resolution images are used in many applications such as, satellite image, medical image etc. In resolution enhancement of images the main loss is high frequency contents (edges) of the image. Hence in order to enhance the quality of image, preserving the edges is necessary. This paper compares various image resolution enhancement techniques such as discrete wavelet transform (DWT), stationary wavelet transform(SWT), dual tree complex wavelet transform (DT-CWT), wavelet zero padding (WZP), cycle spinning (CS), Vector-Valued Image Regularization with Partial Differential Equations (VVIR-PDE), Inter Sub band Correlation Technique (ISC).

View PDFchevron_right
Different Implemented Techniques of Super Resolution Imaging
Joy Acharya

2015

Resolution plays a major role for interpretation and analysis of an image. Super Resolution is a technique to enhance the resolution of an image from single or multiple low resolved images, which gives detailed information present in an image. In this paper, we describe several methods for Super Resolution (SR) that enhances the quality of an image. Mainly the methods are divided into frequency domain and spatial domains. Here, we stated comparison of different approaches, challenges and issues for SR and applications of SR in practical world e.g. in medical imaging, satellite imaging, and forensics. We have approached SR using learning based techniques. We present a novel self-learning approach with multiple kernel learning for adaptive kernel selection for SR. The Multiple Kernel Learning is theoretically and technically very attractive, because it learns the kernel weights and the classifier simultaneously based on the margin criterion. With theoretical supports of kernel matching search method and Optimization approach (Gradient) are proposed our SR framework learns and selects the optimal Kernel ridge regression model when producing an SR image, which results in the minimum SR reconstruction error.

View PDFchevron_right
Analysis of Multi Frame Super Resolution Methods
Editor IJAERD

The matter of Resolution Enhancement has become one of the most important Electronic and Dig ital processing applications in the last few decades. Super Resolution is one of the techniques used for Resolution Enhancement. Super resolution techniques are used to provide higher resolution of an image without distorting an image. Super resolution can be done using two ways: Single frame and Multi frame. For the former directly an interpolation technique can be applied to get a high resolution image fro m a lo w resolution image and for later mu ltiple low resolutions images can be combined to form a h igh resolution image. Using this techniques resolution of any digital or electronic system can be enhanced. This paper addresses the matter of image super-resolution, which is the strategy by which one or more lo w-resolution images are processed together to create an image of high quality or set of images with a higher spatial resolution. Multi-frame image super-resolution refers to the particular case where mu ltip le images of the spot are available. Iterati ve B ack projecti on, Projecti on onto convex sets (POCS)

View PDFchevron_right
Multi-frame super-resolution: A survey
Mahmoud Khattab

IEEE, 2018

One of the primary measurements of image quality is image resolution. High-resolution images are often required and desired for most of applications as they embody supplementary information. However, the best utilization of image sensors and optical technologies to increase the image pixel density is usually restrictive and overpriced. Therefore, the effective use of image processing techniques for acquiring a high-resolution image generated from low-resolution images is an inexpensive and powerful solution. This kind of image improvement is named image super-resolution. This paper undertakes to investigate the current super-resolution approaches adopted to generate a highresolution image. Furthermore, it highlights the strengths and the limitations of these approaches. More to the point, several image quality metrics are discussed to measure the similarity between the reconstructed image and the original image.

View PDFchevron_right
IMAGE RECONSTRUCTION WITH SUPER RESOLUTION
IJRCAR JOURNAL

Subject identification from surveillance imagery has become an important task for forensic investigation. Good quality images of the subjects are essential for the surveillance footage to be useful. However, surveillance videos are of low resolution due to data storage requirements. In addition, subjects typically occupy a small portion of a camera's field of view. Faces, which are of primary interest, occupy an even smaller array of pixels. For reliable face recognition from surveillance video, there is a need to generate higher resolution images of the subject's face from low-resolution video. Super-resolution image reconstruction is a signal processing based approach that aims to reconstruct a high-resolution image by combining a number of low-resolution images. Super-resolution imaging (SR) is a class of techniques that enhance the resolution of an imaging system. Super-resolution image reconstruction is a signal processing based approach that aims to reconstruct a high-resolution image by combining a number of low-resolution images. Super-resolution improves image fidelity, and hence should improve the ability to distinguish between faces and consequently automatic face recognition accuracy.

View PDFchevron_right

References (27)

  1. Sung Cheol Park, Min Kyn Park and Moon Gi Kang, "Super- Resolution Image Reconstruction: A Technical Overview", IEEE signal processing magazine, pp 21-36, May 2003.
  2. Aggelos K. Katsaggelos, Rafael Molina, Javier Mateos, "Super Resolution of Images and Video", Morgan and Claypool Publishers, 1/e, 2007.
  3. Jonathan Cheung-Wai Chan, Jianglin Ma, Frank Canters, "A comparison of super resolution reconstruction methods for multiangle CHRIS/Probe images" Proceedings of spie, image and signal processing for remote sensing xiv, September 15-18, 2008, university of wales institute, Cardiff, Wales, United Kingdom. Vol. 7109, paper no. 710904, pp 1-11. Doi: 10.1117/12.800256.
  4. Subhasis Chaudhuri, "super-resolution imaging", Kluwer Academic Publishers, 1/e, 2002.
  5. Sean borman, robert l. Stevenson "super-resolution from image sequences -a review" Midwest Symposium on Circuits and Systems, 1998.
  6. Sina Farsiu, M. Dirk Robinson, Michael Elad, and Peyman Milanfar, "Fast and Robust Multiframe Super Resolution", IEEE transactions on image processing, VOL. 13, NO. 10, pp 1327-1344, Oct. 2004.
  7. David Capel and Andrew Zisserman, "Super-resolution Enhancement of Text Image Sequences", IEEE Comput. Soc, Proceedings 15th International Conference on Pattern Recognition ICPR, VOL. 1, No. 1, pp-600-605,2000.
  8. Xie Qinlan, Chen Hong, Cao Huimin, "Improved Example-based Single-image Superresolution", IEEE CISP, pp.1204-1207, 2010.
  9. Changhyun Kim, Kyuha Choi, and Jong Beom Ra, "Improvement On Learning-Based Super-Resolution By Adopting Residual Information And Patch Reliability", IEEE ICIP, pp.1197-1200,2009.
  10. Changhyun Kim, Kyuha Choi, Kyuyoung Hwang, and Jong Beom Ra, "learning-based super-resolution using a multi-resolution wavelet approach" Proceedings of 2009 Iternational workshop on Advance Image Technology IWAIT,Jan,2009.
  11. Liyakathunisa, and V. K. Ananthashayana, "Super Resolution Blind Reconstruction of Low Resolution Images using Wavelets based Fusion", World Academy of Science, Engineering and Technology 40, pp 177-181, 2008.
  12. H. Zheng, A. Bouzerdoum, and S. L. Phung, "wavelet based nonlocal- means superresolution for video sequences", Proceedings of 2010 IEEE 17th International Conference on Image Processing, Hong Kong, pp 2817-2820, Sept. 26-29, 2010.
  13. Vandewalle, P., Susstrunk, S., and Vetterli, M. (2003) "Superresolution images reconstructed from aliased images", In Ebrahimi, T. and Sikora, T. (eds.), SPIE Visual Communication and Image Processing Conference, Lugano, Switzerland, pp. 13981405, July.
  14. Priyam Chatterjee, Sujata Mukherjee, Subhasis Chaudhuri and Guna Seetharaman "Application of Papoulis-Gerchberg Method in Image Super-resolution and Inpainting", The Computer Journal Vol. 00 No. 0, 2007.
  15. M. Irani and S. Peleg, "Super resolution from image sequence", in Proceedings of 10 th International Conference on Pattern Recognition (ICPR), pp.115-120, 1990.
  16. K. A. T. Komatsu, T. Igarashi and T. Saito, "Very high resolution imaging scheme with multiple different-aperture cameras", Signal Processing: Image Communication, vol. 5, no. 5-6, pp. 511 526, 1993, special Issue on Digital High Definition Television.
  17. Marcia L. S. Aguena, Nelson D. A. Mascarenhas, "Generalization of Iterative RestorationTechniques for Super-Resolution",SIBGRAPI Conference on Graphics, patterns and Images, pp. 258-265, Aug.2011.
  18. S. P. Kim and W. Su, "Recursive high-resolution reconstruction of blurred multiframe images", IEEE Transactions on Image Processing, vol. 2, pp. 534539, 1993.
  19. Jianchao Yang, John Wright, Thomas Huang and Yi Ma, "Image Super-Resolution via Sparse Representation", IEEE transactions on image processing, vol. 19, no. 11, 2861-1873, Nov. 2010.
  20. Jianchao Yang, John Wright, Yi Ma, Thomas Huang, "Image Super- Resolution as Sparse Representation of Raw Image Patches", IEEE Conference on Computer Vision and Pattern Recongnition (CVPR),pp.1-8,2008.
  21. Yongqiang Zhao, Jinxiang Yang, Qingyong Zhang, Lin Song, Yongmei Cheng and Quan Pan, "Hyperspectral imagery super- resolution by sparse representation and spectral regularization", EURASIP Journal on Advances in Signal Processing, 2011.
  22. Eran Gur, IAENG and Zeev Zalevsky, "Single-Image Digital Super- Resolution A Revised Gerchberg-Papoulis Algorithm", IAENG International Journal of Computer Science, vol. 34, no. 2.
  23. Weisheng Dong, Guangming Shi, Lei Zhang, and Xiaolin Wu, "Super- resolution with Nonlocal Regularized Sparse Representation", Proc. of SPIE Vol. 7744 pp 77440H-1 -77440H-10.
  24. J. T. Hsu, C. C. Yen, C. C. Li, M. Sun, B. Tian3 and M. Kaygusuz, "Application of Wavelet-based POCS Supe rresolution for Cardiovascular MRI Image Enhancement",IEEE Proceedings of the Third International Conference on Image and Graphics, pp 572-575, 2004.
  25. Xiaoqing Su, Shutao Li, "Multi-frame Image Super-resolution Reconstruction based on Sparse Representation and POCS", International Journal of Digital Content Technology and its Applications. Volume 5, Number 8, pp 127-135, Aug. 2011.
  26. Hong Yu, Ma Xiang, Huang Hua, Qi Chun, "Face Image Super- resolution Through POCS and Residue Compensation",VIE, pp 494- 497, 2008.
  27. Frederick W. Wheeler, Ralph T. Hoctor and Eamon B. Barrett, "Super-Resolution Image Synthesis using Projections onto Convex Sets in the Frequency Domain", SPIE 2005.

Related papers

A Review on Super Resolution Techniques
International Journal of Advance Research in Computer Science and Management Studies [IJARCSMS] ijarcsms.com

In review paper [4], authors Huahua Chen, Baolin Jiang, Weiqiang Chen have demonstrated that a super-resolution based on image patches structure. This method have not only has better quality but less consuming time than Yang [11] method.

View PDFchevron_right
A Review on Super Resolution Technique
IJSTE - International Journal of Science Technology and Engineering

In this paper, we summary about the different research papers that applicable to our topic of Research which mentioned above. Image Super Resolution is a most important subject of research in the area of Image Processing. The "super resolution image" refers as technique to produce the high resolution of image from single or multiple low resolution images. The basically idea for Super-Resolution (SR) is that the fusion of a sequence of low-resolution (LR) which are noisy and blurred images that create a higher resolution (HR) image. For super resolution which is process of restoring and de noising of image. Types of resolution methods have been used so far can be divided into three groups as frequency-domain methods, spatial domain methods and techniques can be classified as the wavelet domain.

View PDFchevron_right
A STUDY ON IMAGE SUPER-RESOLUTION TECHNIQUES
IAEME Publication

Image Super-Resolution (SR) is a technique to reconstruct High-Resolution (HR) images using one or more Low-Resolution (LR) images. This paper brings about a detailed study on image Super-Resolution Techniques. Different categories of image Resolution and the process, Image Super-Resolution are well described. A detailed description of different SR approaches is given and certain relevant SR methods are explained. This paper also gives a qualitative and quantitative performance evaluation and comparison of various SR methods.

View PDFchevron_right
Survey of Super Resolution Techniques
Anand Deshpande

ICTACT Journal on Image and Video Processing, 2019

The image processing field is quite advance now a days. Many important developments have taken place over the last three or four decades. Super Resolution is really very important subject for Image Processing. Super resolution describes increasing the resolution of an image using various algorithms and construct a high resolution image from one or more low resolution images. This paper reviews various superresolution technique with their advantages and disadvantages. Finally presented challenge issues and future research directions for super resolution.

View PDFchevron_right
Super Resolution Applications in Modern Digital Image Processing
JAG R O O P SINGH

International Journal of Computer Applications, 2016

This paper gives the various applications of Super Resolution (SR) in today's time modern digital image processing. It has been seen that SR implementation results are always increasing the performance of digital image processing techniques. Modern digital communication includes many applications which generate the data in the form of images and videos. Along with the other techniques of processing of images and videos, Super resolution of images has also grown. In the world of powerful processors and advanced display systems, SR has shown its existence and an overview of various applications based on super resolution has been given in this paper.

View PDFchevron_right
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved