Jump to content

Word search within pictures saved locally?

Recommended Posts


I wonder if  EverNote can do the following: I want to get rid of a lot of papers – and want to take a ‘picture’ (.jpeg) of all incoming documents like bank account statements, insurance policies and all other ‘papers’ received by ‘paper-post’ - with the purpose tobe able to search in all the ‘pictures’ of the papers. The requirements are:

  1. Scan all my documents – let’s say 5 000-10 000 pictures (or pages) – using my Samsung mobile and initially save the ‘pictures’ to my mobile memory SD card – and manually copy the jpeg files to my Windows C-drive let’s say once a month.
  2. After the pictures are saved on my Windows C-drive  - I want to search for all documents with the word ‘statement’ => then for example all bank and insurance statements where the word ‘statement’ is written will be displayed (or all ‘pictures’ where the word ‘statement’ is written) – and I can open the picture and see the whole ‘document’ as a whole.
  3. If I want to search for 2 words:  ‘statement’ and ‘BankABC’ => then only documents (or pictures) where both 2 words are included will be found.

    Is this possible with EverNote? I’m really looking forward to fewer papers – and don’t mind saving 20+GB or so locally with jpeg-files instead of 20kg of papers! J


Link to comment
  • Level 5

I would suggest that you consider saving the statements and other text-based documents as PDF files.

In my experience, the search capability is far superior with PDF vs JPG.


#1 - I go one step further and create searchable PDFs during the scanning process.  It takes a few seconds longer, but I get more consistent search results. And I would feed them directly into Evernote from my scanner.


#2 - "See whole document as a whole" Can't be done with JPG if it is a multiple page document. If you have a 4-page Account Statement you will end up with 4 individual JPGs. And added bonus with PDF, you can store multiple pages in a single PDF. 


#3 - That is correct. I would also recommend you develop a consistent naming convention for your titles.

Example - I start with date code, then geography, then specifics, then who

20151031 Maine Bangor - Sams Club - membership renewal for JLB


Also take advantage of tags. It will make your searches more accurate.


There are a lot of posts on this forum regarding going paperless.

Link to comment

Thanks jbenson2 for the reply.


My take is that Evernote cannot handle text in pictures then – and I have to look for alternatives.


The reason is that creating a .pdf files takes longer time compared to just shoot a picture from a mobile camera. (Unless I can create .pdf files from .jpeg using some other software – for example just mark 50 jpeg files – and automatically create 50 .pdf files or something.


There are two practical use cases:

  1. When I come home and open the ‘paper-post’  => I read it and at the same time also take a picture of the page using my mobile and then through the physical paper away. I don’t want to ‘walk away’ to start my ‘home-printer’ and spend 5 minutes starting it up and create a .pdf. It is not a long term solution. 
  2. However - If I receive a paper document with 22 pages – then I want to create a .pdf with my ‘home-printer - but do it with several documents at the same time - for example once a week.

=> I then want to store all jpeg and .pdf files in one folder locally (or maybe several folders ) – and then just search for the words I’m after. I’m after a very ‘low-admin-procedure’.


When you say that …”the search capability is far superior with PDF vs JPG”… Does it mean that Evernote sometimes fails to find text even in good quality pictures from documents? Pls note that I’m only after searching for text based on very good quality ‘physical-papers’ sent from companies etc. my post-mail.


Thanks, JamesB

Link to comment
  • Level 5*

Hi - I'd agree with @jbenson2 that PDF files are more effective than JPEGs for storing documents.  Evernote can and does parse pictures to identify text content,  but regardless of the quality of the document,  you're taking pics with a mobile camera in variable light conditions.  Good text recognition requires ample lighting of one color - either all electric or all natural - and the document needs to be sitting on a contrasting background;  all of which is catered for in a scanner.  Plus scanned documents can be OCR'd - where the pictures of the text are replaced by the actual characters.  This gives smaller files and more accurate searches.  In a JPG,  by comparison,  the text recognition engine will show possible interpretations of a pictured word - so 'house' would also have house/horse/hearse/mouse/louse associations.  A search for 'house' will (probably) find the right picture,  but then so will searches for the other variations...


On your query - Adobe Acrobat (and presumably other PDF editors) will create PDF files from multiple JPGs.


Like JB2 I normally scan paper into my Evernote,  putting all relevant information into the title.  I open the post,  discard the junk,  and stack the items requiring attention or archive.  Since I write stuff,  I have to sit down at the laptop to do some work,  which is when the post get scanned in.  If I'm travelling I'll deal with the content and make notes on the paperwork,  all to be scanned in when I get back to base.  I do scan some paperwork on the move - used to use CamScanner,  but there's now Scannable (pictures to PDF in iOS) or the Evernote document camera (Android and iOS) which does the same.  When I take pictures I use the same naming conventions as for scans,  so the detail is in the picture,  but dates / document types like 'receipt' and headline content like 'replacement tyre' are in the title.

Link to comment
  • Level 5

When you say that …”the search capability is far superior with PDF vs JPG”… Does it mean that Evernote sometimes fails to find text even in good quality pictures from documents? Pls note that I’m only after searching for text based on very good quality ‘physical-papers’ sent from companies etc. my post-mail.


I have a few photos and images in the JPG format. I noticed Evernote finds "imaginary words" that do not exist in the image. It got frustrating, so I add an arbitrary tag "X" to the JPG images. I added -tag:x to my search. This lets me avoid searching those images.


And if you put in a map into Evernote, it will work overtime analyzing every city, town, river and mountain words.


My computer is left on all day, so it does not take much time to scan.  I use a very small portable Fujitsu ScanSnap scanner that is next to my computer for the mail documents. When I moved from Minnesota to Maine last year, I scanned hundreds of old documents. Electronic data weighs a lot less than the paper.



I also use the camera in my Android phone to take PDF's of my thermal receipts with an app called CamScanner.  The app crops right to the edges of the receipts and documents, so there is no background in the PDF.

Link to comment


This topic is now archived and is closed to further replies.

  • Create New...