A New No-Reference Image Quality Measure for Blurred Images in Spatial Domain

De, Kajal; Masilamani, V.

doi:10.12720/joig.1.1.39-42

Cited by 21 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These relations can be used to justify the computational advantage in considering (14) instead of (12) in the SIFT applications. As an example, for d = 4, the cases σ cascade < 8/3 and 8/3 ≤ σ cascade < √ 8 occur respectively two times per octave, giving a reduction in number of operations for 4 filters out of 5 (per octave).…”

Section: Proposed Method: Fast Eboxmentioning

confidence: 99%

See 1 more Smart Citation

Fast and Accurate Gaussian Pyramid Construction by Extended Box Filtering

Konlambigue

Pothin

Honeiné

et al. 2018

2018 26th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Gaussian Pyramid (GP) is one of the most important representations in computer vision. However, the computation of GP is still challenging for real-time applications. In this paper, we propose a novel approach by investigating the extended box filters for an efficient Gaussian approximation. Taking advantages of the cascade configuration, tiny kernels and memory cache, we develop a fast and suitable algorithm for embedded systems, typically smartphones. Experiments with Android NDK show a 5x speed up compared to an optimized CPU-version of the Gaussian smoothing.

show abstract

Section: Proposed Method: Fast Eboxmentioning

confidence: 99%

“…Box filters introduce noise on edges, known in image processing as ringing artifacts. To quantify this, we used the Blur Measure (BM), defined in [14] as :…”

Section: Accuracy Of the Iterative Ebox Filteringmentioning

confidence: 99%

Fast and Accurate Gaussian Pyramid Construction by Extended Box Filtering

Konlambigue

Pothin

Honeiné

et al. 2018

2018 26th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

show abstract

“…The pH sensors were obtained from six different batches. Since the preliminary studies demonstrated a strong agreement between the colour features of 3 to 5 consecutive images of the sensors, one image was captured from each sensor unless the image was blurry [ 4 ] due to ambient vibrations. The images had a size of 2048 by 1536 pixels.…”

Section: Methodsmentioning

confidence: 99%

Smartphone-based quantitative measurements on holographic sensors

Moghaddam

Lowe

2017

PLoS ONE

View full text Add to dashboard Cite

The research reported herein integrates a generic holographic sensor platform and a smartphone-based colour quantification algorithm in order to standardise and improve the determination of the concentration of analytes of interest. The utility of this approach has been exemplified by analysing the replay colour of the captured image of a holographic pH sensor in near real-time. Personalised image encryption followed by a wavelet-based image compression method were applied to secure the image transfer across a bandwidth-limited network to the cloud. The decrypted and decompressed image was processed through four principal steps: Recognition of the hologram in the image with a complex background using a template-based approach, conversion of device-dependent RGB values to device-independent CIEXYZ values using a polynomial model of the camera and computation of the CIEL*a*b* values, use of the colour coordinates of the captured image to segment the image, select the appropriate colour descriptors and, ultimately, locate the region of interest (ROI), i.e. the hologram in this case, and finally, application of a machine learning-based algorithm to correlate the colour coordinates of the ROI to the analyte concentration. Integrating holographic sensors and the colour image processing algorithm potentially offers a cost-effective platform for the remote monitoring of analytes in real time in readily accessible body fluids by minimally trained individuals.

show abstract

“…The blur measurement method [4] mainly involves using Sobel operators to measure image edges, and the measurement results are compared with those obtained using the just noticeable blur (JNB) method. This method yields higher difference of mean opinion scores.…”

Section: Related Workmentioning

confidence: 99%

“…Image distortion-related studies, such as those on blur, the blocking effect, and sharpness, have produced favorable results [5,[12][13]. The blur measure (BM) method [4] is an algorithm based on the NR image quality assessment method, in which the Sobel algorithm is used to detect image edges. In the Sobel measurement method, one-dimensional analysis is first used to identify the pixel with the largest change in edge color, after which two-dimensional analysis is conducted.…”

Section: Introductionmentioning

confidence: 99%

No-Reference Image Quality Assessment Based on Edges

Chang¹,

Chang

2018

AIVP

View full text Add to dashboard Cite

Image quality assessment is a crucial topic in the field of image processing. In this paper, we propose an edge-based no-reference image quality assessment method. The following factor is applied to assess image quality, namely, improved blur measurement. In the improved blur measurement method, we propose an algorithm that improves the accuracy in measuring image blurs and attains effective execution speed in time complexity. Experimental results reveal that using the proposed approach helps attain satisfactory image quality assessment results.

show abstract

A New No-Reference Image Quality Measure for Blurred Images in Spatial Domain

Cited by 21 publications

References 14 publications

Fast and Accurate Gaussian Pyramid Construction by Extended Box Filtering

Fast and Accurate Gaussian Pyramid Construction by Extended Box Filtering

Smartphone-based quantitative measurements on holographic sensors

No-Reference Image Quality Assessment Based on Edges

Contact Info

Product

Resources

About