A mathematical analysis of the DCT coefficient distributions for images

By Edmund Y. Lam, Joseph W. Goodman. IEEE Transactions on Image Processing, 2000

779 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/e41ffe0a1462eb0117e56c1570e3cf7e1bc8b5eb

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Photo Forensics From Rounding Artifacts (2020, background): “Because the AC-coeicients of natural images are zero-mean [27], the negative and positive biases described above will lead to non-zero-mean distributions which in turn will lead to perceptual artifacts in the average 8 × 8 DCT/intensity block.”
An information fidelity criterion for image quality assessment using natural scene statistics (2005, methodology): “NSS have been explicitly incorporated into a number of image processing algorithms: in compression algorithms [22]–[25], denoising algorithms [26]–[28], image modeling [29], image segmentation [30], and texture analysis and synthesis [31].”
Perceptual Video Coding Based on SSIM-Inspired Divisive Normalization (2013, background): “4 shows that the distribution of the normalized transform coefficients in different sub-bands have similar shape but different widths [53], [54], thus their optimal Lagrange multipliers should be different.”
No-reference quality assessment using natural scene statistics: JPEG2000 (2005, methodology): “NSS have been explicitly incorporated into a number of image processing algorithms: in compression algorithms [10] –[12], [25], denoising algorithms [23], [34], [35], image modeling [36], image segmentation [37], and texture analysis and synthesis [38].”
Image information and visual quality (2006, methodology): “NSS have been explicitly incorporated into a number of image processing algorithms: in compression algorithms [21]–[24], denoising algorithms [25]–[27], image modeling [28], image segmentation [29], and texture analysis and synthesis [30].”
From learning models of natural image patches to whole image restoration (2011, background): “There are many explanations to why this non-Gaussian distribution arises in natural images [21, 22, 18].”