How to use VeryPDF Cloud API OCR SDK for automation and getting coordinate (position) of text on image

hi,

"Greetings of the day"

We are developing a web application that require OCR SDK for get position of particular text on the image.
We also require pdf to image conversion for the same..

So my Query is ...
if we upload a pdf and send it to you for conversion then what we will get in return, Will you provide converted images or you will provide only URL for these converted images..

Our Requirement ...
We want converted images on our server for further process also we want your OCR SDK for automation and getting co-ordinate(position) of text on image.

Please provide details on these matter as soon as possible..

Thanks & Regards,
Customer
---------------------------------------------------
>>if we upload a pdf and send it to you for conversion then what we will get in return, Will you provide converted images or you will provide only URL for these converted images..

We will return the URLs for these converted images, you can download these images to your local server from your code easily.

>>Our Requirement ...
>>We want converted images on our server for further process also we want your
>>OCR SDK for automation and getting co-ordinate(position) of text on image.
>>Please provide details on these matter as soon as possible..

Yes, this is possible, you can subscribe "VeryPDF Cloud API Platform" from following web page,

http://www.verypdf.com/online/cloud-api/try-and-buy.html

after you subscribed "VeryPDF Cloud API Platform", please send to us your Order ID, we will send the detailed instructions to you asap.

Just for test purpose, you can open following URL in web browser,

http://online.verypdf.com/api/?apikey=XXXXXXXXXXXXX&app=ocr
&infile=http://online.verypdf.com/examples/cloud-api/multipage.tif&outfile=out&lang=eng&format

You will get following return,

http://online.verypdf.com/u/public/api/20140619-025657-1646062729-out.html

These HTML contents are contain "X, Y, Width, Height" coordinates for each word and each line, you can write a simple application to parse these HTML contents easily,

<div class='ocr_page' id='page_1' title='image "20140619-025657-4646439576.tif"; bbox 0 0 2000 2388; ppageno 0'>
<div class='ocr_carea' id='block_1_1' title="bbox 22 28 1013 87">
<p class='ocr_par' dir='ltr' id='par_1' title="bbox 23 30 1012 83">
<span class='ocr_line' id='line_1' title="bbox 23 30 1012 83">
<span class='ocrx_word' id='word_1' title="bbox 23 32 262 74">Universal</span>
<span class='ocrx_word' id='word_2' title="bbox 278 31 569 73">Declaration</span> <span class='ocrx_word' id='word_3' title="bbox 586 31 637 73">of</span>
<span class='ocrx_word' id='word_4' title="bbox 649 31 836 72">Human</span>
<span class='ocrx_word' id='word_5' title="bbox 853 30 1012 83">Rights</span>
</span>
</p>
</div>
<div class='ocr_carea' id='block_13_13' title="bbox 24 2434 1944 2587">
<p class='ocr_par' dir='ltr' id='par_13' title="bbox 24 2435 1944 2587">
<span class='ocr_line' id='line_34' title="bbox 24 2435 1901 2482">
<span class='ocrx_word' id='word_391' title="bbox 24 2437 187 2482">Nothing</span> <span class='ocrx_word' id='word_392' title="bbox 202 2436 238 2471">in</span> <span class='ocrx_word' id='word_393' title="bbox 253 2436 323 2471">this</span>
<span class='ocrx_word' id='word_394' title="bbox 338 2436 569 2471">Declaration</span>
<span class='ocrx_word' id='word_395' title="bbox 583 2447 668 2481">may</span>
<span class='ocrx_word' id='word_396' title="bbox 682 2436 727 2471">be</span>
<span class='ocrx_word' id='word_397' title="bbox 742 2435 958 2481">interpreted</span>
<span class='ocrx_word' id='word_398' title="bbox 973 2446 1011 2470">as</span>
<span class='ocrx_word' id='word_399' title="bbox 1027 2435 1206 2481">implying</span>
<span class='ocrx_word' id='word_400' title="bbox 1221 2435 1277 2470">for</span>
<span class='ocrx_word' id='word_401' title="bbox 1292 2446 1362 2480">any</span>
<span class='ocrx_word' id='word_402' title="bbox 1377 2435 1485 2475">State,</span>
<span class='ocrx_word' id='word_403' title="bbox 1502 2445 1616 2480">group</span>
<span class='ocrx_word' id='word_404' title="bbox 1631 2445 1672 2469">or</span>
<span class='ocrx_word' id='word_405' title="bbox 1684 2445 1816 2480">person</span>
<span class='ocrx_word' id='word_406' title="bbox 1832 2445 1901 2479">any</span>
</span>
<span class='ocr_line' id='line_35' title="bbox 24 2492 1944 2539">
<span class='ocrx_word' id='word_407' title="bbox 24 2494 118 2539">right</span>
<span class='ocrx_word' id='word_408' title="bbox 132 2500 169 2529">to</span>
<span class='ocrx_word' id='word_409' title="bbox 183 2505 323 2539">engage</span>
<span class='ocrx_word' id='word_410' title="bbox 337 2494 374 2528">in</span>
<span class='ocrx_word' id='word_411' title="bbox 390 2505 459 2539">any</span>
<span class='ocrx_word' id='word_412' title="bbox 474 2494 622 2539">activity</span>
<span class='ocrx_word' id='word_413' title="bbox 636 2505 677 2528">or</span>
<span class='ocrx_word' id='word_414' title="bbox 690 2499 727 2529">to</span>
<span class='ocrx_word' id='word_415' title="bbox 741 2493 903 2539">perform</span>
<span class='ocrx_word' id='word_416' title="bbox 918 2504 988 2538">any</span>
<span class='ocrx_word' id='word_417' title="bbox 1003 2498 1058 2528">act</span>
<span class='ocrx_word' id='word_418' title="bbox 1074 2493 1193 2528">aimed</span>
<span class='ocrx_word' id='word_419' title="bbox 1208 2498 1242 2527">at</span>
<span class='ocrx_word' id='word_420' title="bbox 1256 2492 1315 2527">the</span>
<span class='ocrx_word' id='word_421' title="bbox 1330 2492 1551 2527">destruction</span>
<span class='ocrx_word' id='word_422' title="bbox 1566 2492 1609 2527">of</span>
<span class='ocrx_word' id='word_423' title="bbox 1621 2503 1690 2537">any</span>
<span class='ocrx_word' id='word_424' title="bbox 1705 2492 1748 2527">of</span>
<span class='ocrx_word' id='word_425' title="bbox 1758 2492 1817 2527">the</span>
<span class='ocrx_word' id='word_426' title="bbox 1832 2492 1944 2537">rights</span>
</span>
<span class='ocr_line' id='line_36' title="bbox 25 2551 619 2587">
<span class='ocrx_word' id='word_427' title="bbox 25 2552 95 2587">and</span>
<span class='ocrx_word' id='word_428' title="bbox 109 2552 293 2587">freedoms</span>
<span class='ocrx_word' id='word_429' title="bbox 309 2557 362 2586">set</span>
<span class='ocrx_word' id='word_430' title="bbox 377 2551 471 2586">forth</span>
<span class='ocrx_word' id='word_431' title="bbox 486 2551 619 2586">herein.</span>
</span>
</p>
</div>
</div>

------------------------------------------------------------------

>>(1). As i already asked that, will you provide converted images after conversion or you will provide only URL for these image?
>>Your reply was: you will provide URL ..
>>Requirement: Its fine but we want images on our server, Is there any process that we can get images on our server with these packages or on extra cost for this.

VeryPDF Cloud API will return the URLs of converted images to you, you can download these images to yourself's server, then you can store these images on your server.

>>(2). As per our next requirement, we want your OCR SDK for getting position of text on images.
>>Your reply was: you will provide a URL that is in html format and contain co-ordinates for all texts in our image.
>>Requirement: Can you provide these co-ordinates and information in a text file or in a json format..?

Yes, no problem, we will return the text contents which contain co-ordinates for all texts in your image, you can parse these text contents to get co-ordinates and words easily.

>>(3). Is there any feature for search with your cloud OCR ??
>>What type of search we require ...
>>We require that when we provide you any text from all the text show in pdf.. so can you provide co-ordinate for particular given text with repetition of that particular(all positions for this text on image, whether it is repeated) text on that image.

Yes, this is already be done in our Cloud API Platform, you can use "PDF Extraction API" function to get co-ordinates for each word from PDF file, you can evaluate this function from following web page,

http://www.verypdf.com/app/pdf-extract-tool/online.html

The output format for text positions is like below,

//Text Positions for each Word
word: x=157.06..188.76 y=18.60..32.55 base=30.17 fontSize=11.52 rot=0 link=00000000 'Home'
word: x=197.88..257.12 y=18.60..32.55 base=30.17 fontSize=11.52 rot=0 link=00000000 'PDF-Tools'
word: x=266.21..287.18 y=18.60..32.55 base=30.17 fontSize=11.52 rot=0 link=00000000 'Doc'
word: x=288.38..323.97 y=18.60..32.55 base=30.17 fontSize=11.52 rot=0 link=00000000 'ument'
word: x=333.65..379.00 y=18.60..32.55 base=30.17 fontSize=11.52 rot=0 link=00000000 'Support'
word: x=65.66..182.76  y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'Advanced'
word: x=190.02..237.43 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'PDF'
word: x=245.12..307.43 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'Tools'
word: x=314.31..432.03 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'Command'
word: x=439.29..488.87 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'Line'
word: x=496.13..550.23 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'User'
word: x=557.23..643.64 y=46.95..72.70 base=68.32 fontSize=21.26 rot=0 link=00000000 'Manual'
word: x=8.87..62.31    y=86.81..100.7 base=98.39 fontSize=11.52 rot=0 link=00000000 'Version:'
word: x=65.64..94.2    y=86.81..100.7 base=98.39 fontSize=11.52 rot=0 link=00000000 'v2.0'
word: x=8.87..79.14    y=117.82..137  base=133.8 fontSize=15.95 rot=0 link=00000000 'Content'
word: x=79.86..133.67  y=155.91..169  base=167.4 fontSize=11.52 rot=0 link=00000000 'Overview'
word: x=79.86..131.13  y=172.74..186  base=184.3 fontSize=11.52 rot=0 link=00000000 'Features'

//Text Positions for each Line
line: x=157.06..379.00 y=18.60..32.55   base=30.17  'Home PDF-Tools Doc ument Support'
line: x= 65.66..643.64 y=46.95..72.70   base=68.32  'Advanced PDF Tools Command Line User Manual'
line: x=  8.87..94.23  y=86.81..100.76  base=98.39  'Version: v2.0'
line: x=  8.87..79.14  y=117.82..137.13 base=133.85 'Content'
line: x= 79.86..133.67 y=155.91..169.86 base=167.49 'Overview'
line: x= 79.86..131.13 y=172.74..186.69 base=184.31 'Features'
line: x= 79.86..203.05 y=189.57..203.52 base=201.15 'Command Line Usage'
line: x=115.36..263.33 y=223.23..237.18 base=234.81 'Input and output PDF file'
line: x=115.36..264.56 y=240.07..254.01 base=251.64 'Show PDF file information'
line: x=115.36..253.03 y=256.90..270.84 base=268.47 'Set PDF file information'

The following web page is the user guide for PDF Extraction Tools software,

http://www.verypdf.com/app/pdf-extract-tool/user-guide.html

VeryPDF

VN:F [1.9.20_1166]
Rating: 10.0/10 (1 vote cast)
VN:F [1.9.20_1166]
Rating: 0 (from 0 votes)
How to use VeryPDF Cloud API OCR SDK for automation and getting coordinate (position) of text on image, 10.0 out of 10 based on 1 rating

Related Posts

This entry was posted in VeryPDF Cloud API and tagged , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *


Verify Code   If you cannot see the CheckCode image,please refresh the page again!