Edge-directed Bicubic Color Image Interpolation

Zhou, Dengwen; Xiaoliu, Shen

doi:10.3724/sp.j.1004.2012.00525

Cited by 4 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Image resolution can be enhanced via various image processing techniques. Past decades have witnessed the evolvement of SR‐based on traditional algorithms (Parker et al., 1983; Zhou, 2010) and machine learning methods (Freeman et al., 2002; Zhang et al., 2012), with limited successes. Recently, the deep learning (DL)‐based approaches are aggressively explored for better visual quality and performance.…”

Section: Introductionmentioning

confidence: 99%

Reconstruction of sub‐mm 3D pavement images using recursive generative adversarial network for faster texture measurement

Guo-long

Wang

Yang

2023

Computer aided Civil Eng

View full text Add to dashboard Cite

It is challenging to collect 3D pavement images with desired resolution for accurate texture measurement at driving speeds with current devices, particularly in the longitudinal direction. This paper presents a novel superresolution technique with recursive generative adversarial network, called Pavement Texture Super Resolution Generative Adversarial Network (PT‐SRGAN), to reconstruct 0.1‐mm pavement 3D image from low‐resolution data for faster texture measurement. With the proposed pseudo‐Laplacian pyramid and an array of learning strategies, the developed PT‐SRGAN reconstructs 0.1‐mm 3D texture images with multiple upscaling factors in longitudinal direction. Combined with the evaluation mask, the proposed method is substantially superior to other methods in terms of three metrics when comparing the image quality of reconstructed 0.1‐mm 3D images against ground truth. The preliminary results indicate that the proposed method enables data collection at driving speeds up to 24 mph to collect 3D pavement images at sub‐mm resolution for faster texture measurement.

show abstract

Section: Introductionmentioning

confidence: 99%

Reconstruction of sub‐mm 3D pavement images using recursive generative adversarial network for faster texture measurement

Guo-long

Wang

Yang

2023

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…Image Interpolation is the easiest and most visualized method in super-resolution image reconstruction methods. In order to increase the visual effect of marginal area in reconstructed image, scholars put forward many interpolation algorithm [4][5][6][7][8][9][10] based on edge detection and the main idea of these methods is try to take the interpolation along but not getting through the direction of edge. Li ect [4] put forward NEDI (New Edge-Directed Interpolation).…”

Section: Introductionmentioning

confidence: 99%

A Novel Image Super-resolution Method Combined with Edge Focusing

Shu¹,

Fang

2016

Proceedings of the 2016 2nd Workshop on Advanced Research and Technology in Industry Applications

View full text Add to dashboard Cite

There are some problems existing in traditional image interpolation algorithms such as blockiness, blurring, etc. To overcome these problems, a novel image super-resolution algorithm based on edge detection and different filters was proposed. The algorithm firstly calculates an initial super-resolution image by Bicubic interpolation method. Then edge focusing is used to determine whether the interpolated point is on the edges. Finally, the intensity value of super-resolution image is achieved by calculating the weighting values of filters for the neighboring region. Experimental results demonstrate that the proposed algorithm makes full use of the edge information and local features to obtain better performance.

show abstract

“…In order to enhance the visual effect of edges, many interpolation algorithms have been proposed based on edge detection [13][14]. The main idea of these algorithms is to interpolate pixels along the direction of the edge rather than the region across the edge.…”

Section: Introductionmentioning

confidence: 99%

“…A new edge-directed interpolation (NEDI) algorithm is designed in [13] to try to improve edge pixel interpolation results However, it requires much matrix inversion and thus is time-consuming. Zhou et al present a colorful image interpolation method [14], where the pixel value is determined by the strength of the edge. This method is simple and fast, but it only detects the edge along diagonal which might degrade the reconstruction of visual effect.…”

Section: Introductionmentioning

confidence: 99%