I have a PDF file containing some tabular data.
http://dl.dropbox.com/u/44235928/sample_rotate-0.pdf
I have to extract the tabular data from it. I have tried following with no success :
Select the text and paste it to notepad/excel-sheet. (I am getting junk characters)
Used save as text from Acrobat Reader. It is also giving junk characters and not the actual text.
Tried ApachePDFBox command line utility to extract text from PDF. It is also giving junk characters instead of real texts.
Finally I am trying a OCR solution. I am converting the pdf file into .tif images using ImageMagick and getting those images processed by tesseract OCR.
The OCR solution is not very accurate though( about 80% words matched ).
I tried changing density and geometry of the image created from PDF to get better results from tesseract OCR.
convert -rotate 90 -geometry 10000 -depth 8 -density 800 sample.pdf img_800_10000.tif;
tesseract img_800_10000.tif img_800_10000.tif nobatch letters;
I am not sure for what kind of image( density, geometry, monochromatic, sharpen boundary etc) would be best suited for the OCR.
Please suggest what could be the best possible parameters(density,geometry,depth etc) for generating images from a PDF file, so that the tesseract accuracy will increase.
I am open to other( non-ocr ) solutions as well.
In this case I recommend to NOT use ImageMagick for the PDF -> TIFF conversion. Instead, use Ghostscript. Two reasons:
Using Ghostscript directly will give you more control over individual parameters of the conversion.
ImageMagick cannot do that particular conversion itself -- it will call Ghostscript as its 'delegate' anyway, but will not allow you to give all the same fine-grained control that your own Ghostscript command will give you.
Most of the text in the table of your sample PDF is extremely small (I guess, only 4 or 5 pt high). This makes it rather difficult to run a successful OCR unless you increase the resolution considerably.
Ghostscript uses -r72 by default for image format output (such as TIFF). Tesseract works best with r=300 or r=400 -- but only for a font size from 10-12 pt or higher. Therefor, to compensate for the small text size you should make Ghostscript using a resolution of at least 1200 DPI when it renders the PDF to the image.
Also, you'll have to rotate the image so the text displays in the normal reading direction (not bottom -> top).
This is the command which I would try first:
gs \
-o sample.tif \
-sDEVICE=tiffg4 \
-r1200 \
-dAutoRotatePages=/PageByPage \
sample_rotate-0.pdf
You may need to play with variations of the -r1200 parameter (higher or lower) for best results.
Since a comment asked "How to define the geometry of an image when using Ghostscript as we do in convert?", here is an answer:
It does not make sense to define geometry (that is image dimensions) and resolution for a raster image created by Ghostscript at the same time.
Once you convert a vector based page of a given dimension (such as PDF) into a raster image (such as the TIFF G4 format) giving a desired resolution (as done in the other answer), you already indirectly and implicitly also did set the dimension:
The original PDF dimension of your sample file sample_rotate-0.pdf is 1008x612 points.
At a resolution of 72 DPI (the default Ghostscript uses if not given directly, or -r72 in the Ghostscript command if given directly) the image dimensions will be 1008x612 pixels.
At a resolution of 720 DPI (-r720 in the Ghostscript command) the image dimensions will be 10080x6120 pixels.
At a resolution of 1440 DPI (-r1440 in the Ghostscript command of my other answer) the image dimensions will be 20160x12240 pixels.
At a resolution of 1200 DPI (-r1200 in the Ghostscript command) the image dimensions will be 16800x10200 pixels.
At resolution of 1000 DPI (-r1000 in the Ghostscript command) the image dimensions will be 14000x8500 pixels.
At a resolution of 120 DPI (-r120 in the Ghostscript command) the image dimensions will be 1680x1020 pixels.
At resolution of 100 DPI (-r100 in the Ghostscript command) the image dimensions will be 1400x850 pixels.
If you absolutely insist to specify the dimension/geometry for the output image on the Ghostscript commandline (rather than the resolution), you can do so by adding -gNNNNxMMMM -dPDFFitPage to the commandline.
There you can find decoded content of your file: https://docs.google.com/open?id=0B1YEM-11PerqSHpnb1RQcnJ4cFk
A absolutely sure the OCR is the best way to read pdf file, but you can try REGEX-ing the native content. It going to be be the hard and long way.
Related
I tried to convert Python Tkinter canvas to pdf. For that I used Ghostscript. Here is the code part,
canvas.postscript(file="tmp.ps",colormode='color')
somecommand = "gswin64c -o output.pdf -sDEVICE=pdfwrite -g57750x62070 - dPDFFitPage tmp.ps"
call(somecommand, shell=True)
The output pdf with large size but the pdf shows canvas GUI cropped and it is in bottom left corner of the pdf.
I want to show complete canvas on pdf.
You've specified -dPDFFitPage, but your input file appears to be PostScript (judging by the '.ps' extension and your question title). PDFFitPage works with PDF input. Even using -dPSFitPage of the simpler -dFitPage will only work if the input PostScript program requests a media size. If it doesn't then the interpreter can't tell what its bounding box is, and so cannot scale it to fit the media.
You've also specified a media size in pixels (-g57750x62070) which is entirely inappropriate when the input and output are vector formats. For what it's worth, you are specifying a fixed media size of (approximately) 80 inches by 86 inches, using the default resolution of 720 dpi.
If all you want to do is turn a PostScript file into a PDF file then the simpler:
gs -sDEVICE=pdfwrite -o out.pdf input.ps
is sufficient.
I am new to ghostscript.
I have a pdf which contains a card. i want to crop that card out.
Currently with the understanding of document i am only able to convert the pdf to image but have no luck in cropping.
Saw every other related question but there are not working for me.
This is code i used in batch file for converting the pdf to image:
"C:\Program Files\gs\gs9.50\bin\gswin64c.exe" -sDEVICE=png16m -r300 -o c:\users\jen\desktop\pdf.png -f "c:\users\jen\desktop\pdf.pdf
pause
now i don't know how to crop with it too ?
i want to crop at certain postition like: Left:28 Top:524 Width:492.3 Height:161
EDIT
I will be using this in firebase functions.
Example PDF file THE_PDF_TO_CROP. I want to cutout the blue area of pdf to image.
You need to set several parameters; Firstly you need to specify the width and height of the output bitmap. You can use either -dDEVICEHEIGHTPOINTS and -dDEVICEWIDTHPOINTS, or alternatively you can specify the output size in pixels using -g<x>x<y> where and are the number of pixels in the x and y directions. Obviously that will vary depending on the resolution. You can't (obviously) use fractional pixels.
If you use -dDEVICEWIDTHPOINTS and -dDEVICEHEIGHTPOINTS then you also need to set -dFIXEDMEDIA to tell the interpreter not to use the media size from the PDF file instead.
So that shoould create an output bitmap of the correct size. If you try rendering your file using just that, you will see that it renders just a portion of the page from the bottom left. So now you need to shift the content around so that the portion you want lies at the bottom left of the media. You can do that by using the PageOffset PostScript operator.
You haven't given any numbers, nor supplied an example file, so lets say (for the sake of example) that you want to render a 1 inch by 2 inch portion of the document. Lets further say that you the part you want rendered starts 2.5 inches from the left edge, and 1.5 inches from the bottom edge.
A suitable command line would be:
gs -sDEVICEWIDTHPOINTS=72 -dDEVICEHEIGHTPOINTS=144 -dFIXEDMEDIA -r300 -sDEVICE=png16m -o out.png -c "<</PageOffset [-180 -108]>> setpagedevice" -f input.pdf
Note that PDF (and PostScript) units are 1/72 inch so 72 = 1 inch, 144 = 2 inches. You need to shift the origin of the page down and left, which is why the values for PageOffset are negative.
If that doesn't work for you I'll need to see your PDF file and you'll need to tell me which version of Ghostscript you are using.
i try to convert an .eps File to .png with ghostscript.
The .eps file has a resolution of 1000x1000 px. But the outfile has big white borders on left and on the bottom side.
gs -dNOPAUSE -dBATCH -r1000x1000 -q -sDEVICE=png256 -dDEVICEWIDTHPOINTS=880 -dDEVICEHEIGHTPOINTS=720 -sOutputFile=infile.png infile.eps
EPS files don't have a resolution, so it cannot possibly have a resolution of 1000x1000, especially not 1000x1000 pixels, because that's not a resolution, its a size.
I very much doubt you want to set the resolution to 1000 dpi and at the same time set a media size of 880 points x720 points. That will result in a .png 12000x10000 pixels. (There are 72 points to the inch, which means you are setting a media of 12x10 inches at 1000 doits per inch)
The correct way to handle an EPS file (which is slightly but importantly different to a PostScript file) is to arrange the scaling yourself.
If the dimensions of the resulting image are not important to you, then you can use -dEPSCrop which will produce an image where the dimensions of the media are taken from the comments in the EPS file.
If you require that the image has specific dimensions then you should use -g to set the media size (in pixels), set -dFIXEDMEDIA and set -dEPSFitPage which will scale the EPS to fit the dimensions of the media.
I found the solution :
-dEPSCrop
Not sure what is causing that without seeing the eps file, but you can trim it off with ImageMagick like this:
convert SomeFile.png -trim result.png
ImageMagick is installed on most Linux distros and is available for OSX, and Windows.
I have image.png in truecolor,
palette.png (N colors, where N>256) or text file, where list RGB color palette.
How to get a picture with this palette?
If I use imagemagick:
convert image.png -remap palette.png remap_image.png
It does not work.
convert image.png -map palette.png remap_image.png
Gives a very bad quality. The image is very noisy. File size is bigger than before.
GIMP gives best quality:
Сonvert image to indexed color > use custom palette
But GIMP is GUI. I need to convert a lot images in the console without running the gimp and X.org.
Using a shared palette across multiple images requires a carefully crafted palette. If you don't take great care when using the palette of a single image across many images, the result will be poor.
This needn't be complicated though. If you have accesss to the GIMP (or other tool) which supports truecolor graphics, you can create a large image and fit all of the smaller images into it, then quantize the image to N colors, then use that palette as the source.
you should be able to closely mimic GIMP's behavior in the console using ImageMagick
Once you've got a truecolor image with all the colors you want to quantize,
# Create an 8-bit png from our source, with a 235-color palette as an example.
convert truecolor_source.png -colors 235 palette.png
# Create an 8-bit png from an arbitrary image and use the palette in palette.png
convert sample.png -map palette.png output.png
There are a number of options for down-sampling colors, like dithering. See the ImageMagickv6 example page for an excellent overview with example pictures and code.
Although I still don't exactly understand what you want to do, your currently most recent comment ("Yes, from RGB to palette will set independently. Need set correct quantity of colors"), it sounds like all you want to do is set a strict limit on the amount of colors of a bunch of images, but they don't need to use the same palette.
In that case, the solution is very simple:
convert sample.png -colors 135 output.png
Try playing with the quantization options if the result isn't to your satisfaction.
If the output image is too large for your liking, you can experiment with the -quality option.
If this still isn't satisfactory, please try to explain your goal in a more detailed manner.
Good luck!
cat photo.png | pngnq -s 1 > photoindexed.png
I tend to get good results with the "-remap" (single imge) or "+remap" (multiple images) functions in combination with "-colors". Read up on those functions here. Note that "with "-remap" you provide IM with the final set of colors you want to use for the image, whether you plan to dither those colors, or just replace the ones with their nearest neighbours.", meaning just remapping/replacing might not look good enough, as colors from the input image are simply replaced by those from the palette image. Some form of dithering will be necessary to distribute pixel color conversion errors throughout the output image, because not all colors in the palette match those of the input image.
I'd suggest you use the "-colors N" option for that. This will reduce your output image color count to a maximum of N. By default ImageMagick uses "-dither Riemersma" for this implicitly when you specify "-colors N". The are also other dithering options available.
I have a number of input images that contain multiple smaller images, all of them in a single row. All the contained images are the same size. So, for example, the image input.png may be 480x48 and contain 10 48x48 images, all in one row.
Using the imagemagick convert tool (or any other tool supplied with the defaul imagemagick suite), I want to write a bash script that takes an input image, the number of images to cut, and then cuts them all into individual images.
The user interaction i can do, but I've not been able to get convert to do the actual cutting. Can anyone suggest something? From reading the manual pages, i think this should work:
convert 'input.png[0x0+48+48]' output.png
but I get an error:
convert: no pixels defined in cache
tb_icons_l.png' #
magick/cache.c/OpenCache/3572.
convert: No IDATs written into file
1.png' #
coders/png.c/PNGErrorHandler/1391.
Any ideas?
I would do it like that:
convert input.png -crop 48x48 +repage +adjoin output_%02d.gif
Read more at Tile Cropping, in ImageMagick documentation.
I believe you've got the position and size swapped. Try:
convert 'input.png[48x48+0+0]' output.png
The third image would be:
convert 'input.png[48x48+96+0]' output.png
Or
convert 'input.png[48x48+0+96]' output.png