OCR Falling short
Currently working on setting up OCR at our work and it is having trouble picking up a lot of numbers on tags. I have the confidence set at 0.1, 0.1 and resolution 2160p using the server model. Currently the test set is 20 images total with only 6 containing the tags.
My secondary question is that if I can successfully get OCR to pick these up, is there a way to have the search bar default to OCR rather than context?
This is for commercial use so overconfidence is key.
14 Replies
:wave: Hey @Ok,
Thanks for reaching out to us. Please carefully read this message and follow the recommended actions. This will help us be more effective in our support effort and leave more time for building Immich :immich:.
References
- Container Logs:
docker compose logs docs
- Container Status: docker ps -a docs
- Reverse Proxy: https://immich.app/docs/administration/reverse-proxy
- Code Formatting https://support.discord.com/hc/en-us/articles/210298617-Markdown-Text-101-Chat-Formatting-Bold-Italic-Underline#h_01GY0DAKGXDEHE263BCAYEGFJA
Checklist
I have...
1. :ballot_box_with_check: verified I'm on the latest release(note that mobile app releases may take some time).
2. :ballot_box_with_check: read applicable release notes.
3. :ballot_box_with_check: reviewed the FAQs for known issues.
4. :ballot_box_with_check: reviewed Github for known issues.
5. :ballot_box_with_check: tried accessing Immich via local ip (without a custom reverse proxy).
6. :ballot_box_with_check: uploaded the relevant information (see below).
7. :ballot_box_with_check: tried an incognito window, disabled extensions, cleared mobile app cache, logged out and back in, different browsers, etc. as applicable
(an item can be marked as "complete" by reacting with the appropriate number)
Information
In order to be able to effectively help you, we need you to provide clear information to show what the problem is. The exact details needed vary per case, but here is a list of things to consider:
- Your docker-compose.yml and .env files.
- Logs from all the containers and their status (see above).
- All the troubleshooting steps you've tried so far.
- Any recent changes you've made to Immich or your system.
- Details about your system (both software/OS and hardware).
- Details about your storage (filesystems, type of disks, output of commands like fdisk -l and df -h).
- The version of the Immich server, mobile app, and other relevant pieces.
- Any other information that you think might be relevant.
Please paste files and logs with proper code formatting, and especially avoid blurry screenshots.
Without the right information we can't work out what the problem is. Help us help you ;)
If this ticket can be closed you can use the /close command, and re-open it later if needed.
Successfully submitted, a tag has been added to inform contributors. :white_check_mark:Would paperless ngx not be a better fit?
Possibly, haven't done any testing with it.
The UI needs to be extremely simple as it will be used by a diverse group of people.
I don't know your application, so I can't comment too much. I'd just suggest choosing the right tool for the job
For sure, basically the end goal is we have a worker take a photo of a bundle of lumber, that bundle also has a tag with a barcode as well as the id number. This photo gets uploaded to a drive and when a customer requests a photo for the bundle we can search the tag id, it matches it to the image and we can send it back.
Current implementation uses google keep which requires a shared account but keeps OCR is really good
Im tinkering with paperless right now but even with adjusted settings its still missing
Interesting. I'd go about implementing the solution differently - but I guess you have a direction you're looking to go
How would you go about it? I joined the company and the keep solution is what they had when I came.
I'd tie it into the ERP system
Either way we would need OCR to identify the id and attach it to the asset
That WOULD be the IDEAL ideal solution
Not really. This is what barcodes are for. I'm pretty sure there's fully fledged software solutions to do this correctly.
That's probably not what you're looking for at the moment
Theoretically could do this by having a custom app that has you scan a barcode then take the image of the bundle
If you know of any that are off the shelf it could definitely be an option
This isn't a field I'm commonly in. I don't think ocr will ever be perfect and you typically want to be able to gather more information than just a picture. Date / time of delivery, customer signature if applicable, etc etc.
I'd probably search for some delivery software, possibly using different words.
Even better would be if the quotation / invoicing software has a plugin available for this
I don't believe it does, the weird part is that OCR via google keep has worked perfectly so far
The only down side is now that the google keep is full it is very slow
But to answer your question, if you want better ocr results the only thing you can do is increase resolution
(in immich, currently at least)
Remember it takes your preview image and resamples from there