Editing Compression artifact (section)

== Images ==
<!-- deleted file removed [[File:Difference 60 x5.jpg|150px|thumb|This image shows the [[Residual (numerical analysis)|residual]] (with amplification) between [[:Image:60 save for webx5.jpg|a JPEG image]] and [[:Image:Original crop.jpg|the original file]]. Note especially the changes apparent on sharp edges.]] -->
[[File:Jpeg-text-artifacts.gif|thumb|right|320px|Illustration of the effect of JPEG compression on a slightly noisy image with a mixture of text and whitespace. Text is a screen capture from a Wikipedia conversation with noise added (intensity 10 in Paint.NET). One frame of the animation was saved as a JPEG (quality 90) and reloaded. Both frames were then zoomed by a factor of 4 (nearest neighbor interpolation).]]

When performing block-based [[discrete cosine transform]] (DCT)<ref name="Katsaggelos"/> coding for [[Quantization (signal processing)|quantization]], as in [[JPEG]]-compressed images, several types of artifacts can appear.

* [[Ringing artifacts#JPEG|Ringing]]
* Contouring{{clarify|date=February 2021}}
* [[Posterizing]]
* Staircase noise ([[aliasing]]) along curving edges
* Blockiness in "busy" regions (block boundary artifacts, sometimes called (macro)blocking, quilting, or checkerboarding)

Other lossy algorithms, which use [[pattern matching]] to deduplicate similar symbols, are prone to introducing hard to detect errors in printed text. For example, the numbers "6" and "8" may get replaced. This has been observed to happen with [[JBIG2]] in certain photocopier machines.<ref name="kriesel">{{cite web|url=http://www.dkriesel.com/en/blog/2013/0802_xerox-workcentres_are_switching_written_numbers_when_scanning|title=Xerox scanners/photocopiers randomly alter numbers in scanned documents|date=2013-08-02|access-date=2013-08-04}}</ref><ref>{{cite news|url=https://www.bbc.co.uk/news/technology-23588202 |title=Confused Xerox copiers rewrite documents, expert finds |work=BBC News |date=2013-08-06 |access-date=2013-08-06}}</ref>

=== Block boundary artifacts ===
[[File:The macroblocking effect (JPEG).png|344px|right|thumb|Block coding artifacts in a JPEG image. Flat blocks are caused by coarse quantization. Discontinuities at transform block boundaries are visible.]]

At low bit rates, any [[lossy]] block-based coding scheme introduces visible artifacts in pixel blocks and at block boundaries. These boundaries can be transform block boundaries, prediction block boundaries, or both, and may coincide with [[macroblock]] boundaries. The term ''macroblocking'' is commonly used regardless of the artifact's cause. Other names include blocking,<ref name="amiriImageCompressionUsing2018">{{Cite journal| doi = 10.1007/s11042-017-4763-1| issn = 1573-7721| volume = 77| issue = 7| pages = 8677–8693| last1 = Amiri| first1 = Sekine Asadi| last2 = Hassanpour| first2 = Hamid| title = Image compression using JPEG with reduced blocking effects via adaptive down-sampling and self-learning image sparse representation| journal = Multimedia Tools and Applications| access-date = 2024-03-08| date = 2018-04-01| url = https://doi.org/10.1007/s11042-017-4763-1| url-access = subscription}}</ref> tiling,<ref>{{cite book|url=https://books.google.com/books?id=STFAx4jqU5IC&q=tiling+mpeg&pg=PA343|title=The MPEG handbook by John Watkinson|isbn=9780240805788|last1=Watkinson|first1=John|year=2004|publisher=Taylor & Francis }}</ref> mosaicing, pixelating, quilting, and checkerboarding.

Block-artifacts are a result of the very principle of [[Transform coding#Digital|block transform]] coding. The transform (for example the discrete cosine transform) is applied to a block of pixels, and to achieve lossy compression, the transform coefficients of each block are [[Quantization (image processing)|quantized]]. The lower the bit rate, the more coarsely the coefficients are represented and the more coefficients are quantized to zero. Statistically, images have more low-[[Spatial frequency|frequency]] than high-frequency content, so it is the low-frequency content that remains after quantization, which results in blurry, low-resolution blocks. In the most extreme case only the DC-coefficient, that is the coefficient which represents the average color of a block, is retained, and the transform block is only a single color after reconstruction.

Because this quantization process is applied individually in each block, neighboring blocks quantize coefficients differently. This leads to discontinuities at the block boundaries. These are most visible in flat areas, where there is little detail to mask the effect.

=== Image artifact reduction ===
{{Main article|Deblocking filter}}

Various approaches have been proposed to reduce image compression effects, but to use standardized compression/decompression techniques and retain the benefits of compression (for instance, lower transmission and storage costs), many of these methods focus on "post-processing"—that is, processing images when received or viewed. No post-processing technique has been shown to improve image quality in all cases; consequently, none has garnered widespread acceptance, though some have been implemented and are in use in proprietary systems. Many photo editing programs, for instance, have proprietary JPEG artifact reduction algorithms built-in. Consumer equipment often calls this post-processing "MPEG Noise Reduction".<ref>{{cite web|url=https://www.pcmag.com/encyclopedia_term/0%2C2542%2Ct%3Dblocking+artifacts%26i%3D56319%2C00.asp|title=PC Magazine, Definition of blocking artifacts|access-date=23 September 2020|archive-date=7 October 2012|archive-url=https://web.archive.org/web/20121007015121/http://www.pcmag.com/encyclopedia_term/0%2C2542%2Ct%3Dblocking+artifacts%26i%3D56319%2C00.asp|url-status=dead}}</ref>

Boundary artifact in JPEG can be turned into more pleasing "grains" not unlike those in high ISO photographic films. Instead of just multiplying the quantized coefficients with the quantisation step {{mvar|Q}} pertaining to the 2D-frequency, intelligent noise in the form of a random number in the interval {{math|1=[-''Q''/2; ''Q''/2]}} can be added to the dequantized coefficient. This method can be added as an integral part to JPEG decompressors working on the trillions of existing and future JPEG images. As such it is not a "post-processing" technique.<ref>{{cite journal |last1=Hudson |first1=Graham |last2=Léger |first2=Alain |last3=Niss |first3=Birger |last4=Sebestyén |first4=István |last5=Vaaben |first5=Jørgen |title=JPEG.1 standard 25 years: past, present and future reasons for success |journal=Journal of Electronic Imaging |date=31 August 2018 |volume=27|issue=4 |page=1 |doi=10.1117/1.JEI.27.4.040901 |s2cid=52164892 |doi-access=free }}</ref>

The ringing issue can be reduced at encode time by overshooting the DCT values, clamping the rings away.<ref>{{cite book |last1=Richter |first1=Thomas |title=2016 IEEE International Conference on Image Processing (ICIP) |chapter=JPEG on STEROIDS: Common optimization techniques for JPEG image compression |date=September 2016 |pages=61–65 |doi=10.1109/ICIP.2016.7532319 |isbn=978-1-4673-9961-6 |s2cid=14922251}}</ref>

Posterization generally only happens at low quality, when the DC values are given too little importance. Tuning the quantization table helps.<ref>{{cite web |title=kornelski/jpeg-compressor |website=[[GitHub]] |url=https://github.com/kornelski/jpeg-compressor#low-dc-quantization |date=16 November 2020}}</ref>