LP-exam Online(Free)



Please enter the source(URL, DOI).


Upload a file.


Push the button.


解析結果カタログ に解析例として掲載させていただく場合があります )。

Permit result to the public.

■使用例  /  Example

STAP細胞に関する電気泳動画像の分析例 / Example of Our Analysis


Detection of image manipulations in an electrophoresis picture found in the previously mentioned “STAP Cell” paper .











※クリックすると拡大します / Click to enlarge


▼上記の画像について / Explanations

(1)原画像(微調整あり) (2)明暗反転画像 (3)カラーマップ画像
(4)~(6)段階的なコントラスト強調(弱(4), 中(5), 強(6))
(7)~(9)段階的な高域通過処理(スケール 小(7), 中(8), 大(9))
(4)~(6)… 切り貼りがされた箇所全体(赤枠内)が目立って見える。
(7)~(9) … 連続的な直線(青枠内:黒色直線)が見えるようになる。
(3) … 一つのバンドが全て同じ色に見えることもある。

*注1) 小保方氏等の Nature 論文 “Stimulus-triggered fate conversion of somatic cells into pluripotency” (http://www.nature.com/nature/journal/v505/n7485/full/nature12968.html) Fig. 1iの画像を使っています。
*注2) 画像不正の詳細な定義
生命科学分野で扱われる画像には撮影装置や光学的現象に由来する様々なノイズが含まれています。そのため、画像に対して一連の画像処理 (コントラスト調整やノイズ除去) を施すことは、実験データを簡潔かつ明瞭に説明するために必要なことがあります。しかし画像処理を施した場合、論文中に「施した画像処理」 および 「使用した画像処理ソフトウェア」 の記述が厳密に求められています。英科学雑誌「ネイチャー」の投稿規定によると、
・ 施した画像処理は論文中で明言する
・ 処理は画像全体に対して行ない、画像の一部分だけに画像処理を施してはならない
・ 実験データが隠れるほどの過大なコントラスト調整を施してはならない
等、画像処理に対する規定が厳密に定められています(http://www.nature.com/authors/policies/image.html) 。多くの科学雑誌は画像処理に対して同様の投稿規定を定めています。したがって、画像中に 「切り貼りされた跡」 があり 「画像処理を施したことが論文中に明言」されていない場合、画像不正に該当すると考えられます。

1. Transformation into a dark background

2. Conversions between negative and positive coloring

3. Color mapping

4 to 6. Step-by-step contrast emphasis (4. Weak, 5. Medium, 6. Strong)

7 to 9. Step-by-step high pass filters (Scaling 7. Small, 8. Middle, 9. Large)

If these images are suspected to be falsified, the features listed below will further prove the image alterations.

4 to 6. The pasted image fragment (equivalent to the area edged with red lines) is emphasized.

7 to 9. An unusual line (the black line edged with blue lines) emerges.

3. The scale of each band may be lost.

As you can observe, images 4. to 9. clearly indicate that the altered images in this paper were merely images acquired from pervious experiments. These images were cut and pasted into the areas bordered by red lines.

According to the editorial guidelines of Nature, there are clear-cut rules for image processing; for example: 
◯Authors should list all image acquisition tools and image processing software packages used. Authors should document key image-gathering settings and processing manipulations in the Methods.
◯The use of touch-up tools, such as cloning and healing tools in Photoshop, or any feature that deliberately obscures manipulations, is to be avoided.
◯Processing (such as changing brightness and contrast) is appropriate only when it is applied equally across the entire image and is applied equally to controls. Contrast should not be adjusted so that data disappear. Excessive manipulations, such as processing to emphasize one region in the image at the expense of others (for example, through the use of a biased choice of threshold settings), is inappropriate, as is emphasizing experimental data relative to the control.
See also (http://www.nature.com/authors/policies/image.html).



▼解析の様子の例(動画)/ Analysis procedures (movie)

>>English ver.


LP-exam による解析結果カタログ / Album



▼LP-exam Ver.1オンライン版

・画像の明暗反転 … 暗色(収縮色)背景と明色(膨張色)背景では画像全体の印象が異なるので、画像の客観的な識別に有効である。
・カラーマップ画像 … 輝度分布の視認性に優れ、モノクロ表示ではわかりにくい明るさの違いが鮮明になる。
・段階的コントラスト強調 … 切り貼りされた箇所は周囲のピクセルと輝度が異なる場合が多いため、段階的にコントラストを調整することで、輝度の変化がより明確に検出できる。
・高域通過処理 … 高域通過処理により、画像上目立たない明るさの変化を強調し浮かびあがらせる。














▼LP-exam online Ver.

○Here are the distinctive algorithms used in the Image-manipulation detection software.
・Conversion between negative and positive colored images … Converting positive colored images into negatives enables the viewer to observe detailed differences between the falsified image and its original copy. For example, if the alteration between the manipulated image and its master copy only varied in colors that are difficult to distinguish in positive coloring, a negative counterpart will display a more apparent difference.

・Color mapping … Color mapping emphasizes the variance of brightness in an image by improving the clarity of its brightness distribution. Without color mapping, it may be difficult to differentiate certain photographic details when looking at a simple black-and-white background.
・Step-by-step contrast emphasis… Contrast emphasis can discriminate the differences in the image’s brightness, making it possible to detect whether or not the image has been merely pasted over another similar image. We encountered many cases where significant variance in the image’s brightness was detected, especially in locations where copy and pasting has occurred.

・High pass filters … High pass filters is a method that detects fine-tuned brightness alterations in an image, which is normally very difficult for the human eye to distinguish.

○Falsification detection

・LP-exam can detect image manipulations without undergoing complicated procedures.

○Compatible file formats

・Compatible file formats are as follows: PNG, BMP, TIFF, JPEG, GIF and PDF.

○An extension of the LP-exam is already underway. In the near future, it will include further powerful functions such as:

・Detection system for copyright violations

・Analytical system that can recognize the specific machinery responsible for taking the images of concern (e.g. MRIs, microscopes, etc.).

・Falsification judgment and screening system for images.

※The offline version of LP-exam (currently not distributed) is capable of carrying out a much higher level of analysis in comparison to the online version.



本サービス(オンライン版)はクラウドファンディングで募った資金で運営しております。 READY FOR?Special thanks for all investors at Ready For?(Cloud Founding)!

【謝辞】※ReadyForの出資者の皆様 (敬称略・五十音順)

【Special Thanks】※Investors of ReadyFor (Cloud founding)

井石 彩
谷垣 洸哉
中島 孝幸
富永 民雄
原 壮太
牧之段 浩平
松尾 豊
Asuka Kuwabara
Atsushi “JIMMY” Kawase
Jun Dobashi
K. A.
Shigenori Nonaka




[User policy] 

・When you use tools on this web site, please consent to the Terms and Conditions below.

・You are required to consent to the Terms and Conditions when use LP-exam(Ver.1).

・LPixel Inc. won’t take any responsibility to results of LP-exam.

・When you use LP-exam, you are expected to take any responsibility to the results. LPixel won’t compensate any losses caused by LP-exam nor results of the software.

・You are required to be given permission by LPixel if you intend to use LP-exam on a

commercial basis.

・In case you publish results of LP-exam, please write “The result(s) of LP-exam from LPixel

Inc.” clearly.

・It is prohibited to use LP-exam for any image analysis that may insult someone, or cause

people to feel disgusted.

・Please be aware that LPixel can take advantage of any information acquired from the

service at all times.

・Updating of the software including addition of new functions, or ceasing of the distribution

may be occurred without previous notice.

・Please inform LPixel of any defect that you came across to better the software.



・佐野 行己(東京大学)/ Yukimi Sano, The University of the Tokyo
・Ira Horecka, University of California, Santa Cruz