Interpolation based on similarity in subband generation for image super resolution

Dhara, Sobhan Kanti; Sen, Debashis

doi:10.1109/tencon.2017.8227890

Cited by 2 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…C Lien et al, [18] planned a learning-based SR technique with wavelet coefficients estimation to reconstruct the HR images. Dhara and Sen [19] offered an interpolated SR method that performs approximations of high frequency components by manipulating similarities in the process of wavelet breakdown of an image. The intended attenuation is achieved by the proper combination of DWT and stationary wavelet transform (SWT).…”

Section: Literature Surveymentioning

confidence: 99%

DWT based Image Super Resolution using Segmentation and Edge Detection

P¹,

Jabeen²

2022

YMER

View full text Add to dashboard Cite

An image super-resolution is the development of creating a high-resolution image from low-resolution images for better image visualization for several recent applications. In this research paper, we propose DWT-based Image Super Resolution using Segmentation and Edge Detection for better picturing of images. The different kinds of colore images are considered and converted into greyscale images and resized to a uniform size of 166x304. The Discrete Wavelet Transform (DWT) is used to get one low-frequency and three high-frequency bands with a size of 83x152. The four bands are fused and dividing each coefficient by two to get Low Resolution (LR) images. Each coefficient of the LR image is inflated into 2x2 coefficients to get High Resolution (HR) image. The HR matrix is segmented into a 3x3 overlapped matrix and an average value is calculated and assigned to the whole HR matrix. The Canny edge detection is used on the original image to get the edge detected image. The pixel values of HR and canny edge detected images are compared and high-value coefficients are considered. The guided filter is applied to the highvalue coefficients matrix to improve the quality of the image by refining the image. The LL band of the DWT matrix is padded with zeros to convert matrix size to 166x304 and Inverse Discrete Wavelet Transform (IDWT) is used on it. The refined and IDWT images are merged by computing the average values to acquire Super Resolution (SR) image. It is noted that the anticipated system results are enhanced compared to current methods. Keywords: DWT, Fusion, Image Processing, Segmentation, Super Resolution

show abstract

Section: Literature Surveymentioning

confidence: 99%

DWT based Image Super Resolution using Segmentation and Edge Detection

P¹,

Jabeen²

2022

YMER

View full text Add to dashboard Cite

show abstract

“…We consider comparisons of our approach based on visual inspection as well as PSNR, SSIM, [43] and FSIM [44] measures with the widely accepted edge directed NEDI [13], nonlocal autoregressive modeling based NARM [10], recent self-exemplar based SR-TSE [29], the latest fuzzy predictive composite scheme (FPCS-LAL) [40], and deep learning based EDSR [33] and SRGAN [34]. A preliminary version of this work has been published in [45], where a few other existing approaches were considered for result analysis. They, in general, perform inferior to the existing approaches considered here for comparison.…”

Section: Introductionmentioning

confidence: 99%

“…and deep learning based EDSR [33] and SRGAN [34]. A preliminary version of this work has been published in [45], where a few other existing approaches were considered for result analysis. They, in general, perform inferior to the existing approaches considered here for comparison.…”

mentioning

confidence: 99%

Across-scale process similarity based interpolation for image super-resolution

Dhara

Sen

2019

Applied Soft Computing

Self Cite

View full text Add to dashboard Cite

A pivotal step in image super-resolution techniques is interpolation, which aims at generating high resolution images without introducing artifacts such as blurring and ringing. In this paper, we propose a technique that performs interpolation through an infusion of high frequency signal components computed by exploiting 'process similarity'. By 'process similarity', we refer to the resemblance between a decomposition of the image at a resolution to the decomposition of the image at another resolution. In our approach, the decompositions generating image details and approximations are obtained through the discrete wavelet (DWT) and stationary wavelet (SWT) transforms. The complementary nature of DWT and SWT is leveraged to get the structural relation between the input image and its low resolution approximation. The structural relation is represented by optimal model parameters obtained through particle swarm optimization (PSO). Owing to process similarity, these parameters are used to generate the high resolution output image from the input image. The proposed approach is compared with six existing techniques qualitatively and in terms of PSNR, SSIM, and FSIM measures, along with computation time (CPU time).It is found that our approach is the fastest in terms of CPU time and produces comparable results. Recently, few end to end deep neural networks have been proposed for superresolution. SRCNN [30, 31] uses the convolutional neural network (CNN) [32] for image super-resolution. The technique is later extended to DRCN [9] by increasing the number of layers of the network and incorporating recursive learning.The techniques proposed in [8,11] use residual learning in deep and very deep If the aforesaid wavelet based analysis and synthesis is considered for interpolation, then, the input low resolution image can be decomposed into its immediate lower resolution approximation (LL) using (2) and high frequency details using (3). Now, one can assume that only the lower approximation is available for synthesizing back the input image, and not the high frequency details. In such a case, one can try to estimate the unavailable high frequency

show abstract

Interpolation based on similarity in subband generation for image super resolution

Cited by 2 publications

References 12 publications

DWT based Image Super Resolution using Segmentation and Edge Detection

DWT based Image Super Resolution using Segmentation and Edge Detection

Across-scale process similarity based interpolation for image super-resolution

Contact Info

Product

Resources

About