View OCR Text Recognised

Hi - am I able to view the OCR text that has been recognised in Mylio? For example, I used the search term “ocr:*” to find all images that have OCR text (please correct me if I’m wrong on this!). One of the images has me drinking a beer with the word “Belhaven” on the glass. If I search for “ocr:belhaven” no results are returned. So hence I’m curious to see what text the OCR actually recognised.
Thanks!

There is a Mylio Console command you can use to accomplish this. Open the View Menu (the three vertical dots in the top-right) and then go to Help > Console. Select a photo (or open it in single-photo view) and run the following console command:

ocr list

Some information will be returned about the selected photo, including the OCR keywords identified.

3 Likes

Also, Mylio will automatically add the “Document” category to any file with OCR in it. So, if you want to narrow down the list of photos that have OCR in them first - before you run the console command - do a filter for Category → Documents. Of course, any file that you have manually categorized as Document will also appear.

2 Likes

Ok, thank you for that. So I selected an image that was returned in my search results from ocr:*. When I run the command ocr list in the console I see the ocrKeywords as expected. A couple of things I noticed though.

  1. The photo is not categorized as Document
  2. When I search for the ocrKeywords shown in the console, the photo is not returned in the search results. So the ocrKeywords are SPEED LIMIT 40, when I search for SPEED or speed that photo is not returned.

Any ideas?

Hmm are you using any other filters or search modifiers? If you can see that OCR Keywords were identified but they aren’t showing up in search results, it’s possible there is an issue with your catalog.

The catalog is simply a file on your devices that holds information about your library, such as how many photos you have and where they are stored. Mylio can use a catalog from one of your other devices to sync a new catalog to the device.

You can try running a catalog repair to see if that resolves things. The steps for the repair process can be found here:

1 Like

So when I do a new search or filter, I first check that both search and filter symbols are white to (hopefully) ensure there are no other filters/searches in place.

I did the catalogue repair procedure on my main local vault (I have a cloud vault as well) and on another PC Mylio instance. I now get the following results:
Total media: 84,083
Media with documents label: 3,020
Media return with ocr:* search: 19,009

Curious if it is expected to have the number of ocr:* search media much greater in number than the number of documents?

This is expected. Media only gets assigned the “Documents” category if it meets certain criteria (at the moment I believe media must have 10 or more words having 5 or more letters).

2 Likes