I see google API has 3 ways of OCR-ing text from an image/document:
detect text
detect handwriting
detect text in files
(1) allows to upload base-64 image, (2) and (3) forces me to only use a file referenced in google cloud?
Can I use (2) and (3) with base-64 instead?
never mind, found it. Google wants you to upload and save in their cloud TIF/PDF so it can be done offline.
Disappointing.
Related
I am having an issue uploading iPhone photos. The default iPhone image is a .HEIC file. Google vision seems to have a hard time reading these file types. It works fine when using .jpg files. Any suggestions?
If you are coding in nodejs maybe use this library to convert and upload:
https://www.npmjs.com/package/heic-convert
There are online tools as well on the web:
https://heictojpg.com/
Heic is not supported directly
You can convert each image and send it to the google API, if you have access to the source iPhone you can change the settings to avoid using Heic as default format.
You can see here relevant question
I would like to use python to download the image and sequences of images found in the location on openstreetcam.
http://openstreetcam.com/details/8552/422
I figured out the image is saved under
http://api.openstreetcam.org/files/photo/2016/6/30/lth/8552_2fbf0_57756eba868e9.jpg?v=1518090956232
however there is no official API to use. How would one extract the image and gps data?
Edit: The GPS data can be found in the url by clicking Edit OSM id.
Ideally one would use some sort of web scraper however the .jpg is not found in the website source code.
OSC's endpoints are a little hidden.
Invite you to check out https://github.com/Streets-Data-Collaborative/osc-tools where I've written some scripts to extract track data and the underlying metadata for each track.
Feel free to open an issue on the repo if something's not working.
I am searching an image via Google image-search with the 'Drag & Drop' method.
My question is, if I drag an image from my computer into Google image-search, it will show the result (if any) but what happens to the dragged (uploaded) image?
Will it be removed/deleted or will it be stored in Google's servers or any another place?
It looks like they will keep a copy of the image.
https://support.google.com/websearch/answer/1325808?p=searchbyimagepage&hl=en
How Google uses the image you search with
When you search by image, any images or URLs that you upload will be
stored by Google. Google uses those images and URLs solely to provide
and improve our products and services.
Is there any chance the option of returning a TIFF image will be a configurable option in the connect profile any time soon? We are working with a few firms that require TIFF for FINRA compliance.
Thanks,
Russell
Upon import into the DocuSign system, all file formats are converted to PDF format -- and will remain in PDF format from that point on. This means that any files you retrieve from DocuSign (via Connect, via API, manually via the web console) will always be in PDF format. This is not likely to change.
I suspect that you'd be able to receive the PDF files via Connect and convert them to TIFF format via some automated process. Although I don't personally have any experience with this (automatically converting PDF to TIFF), a quick Google search turns up lots of possibilities...for example: Best way to convert pdf files to tiff files.
Is it possible to pass a secure KML file to Google Earth?
This KML file will require either a token or authentication of some sort.
(I'll adapt the solution based on the possible methods of securing the KML file that are acceptable.)
We need to display some secure information on a google map, and yes we are an enterprise customer. (If that helps.)
That alternative is to plot the points using the google API, but that require's slightly more effort.
If your intention is just to protect the Kml from unauthorized access, you can use the following approach. I tried this approach in one of the scenarios and it worked for me.
http://keelypavan.blogspot.com/2011/07/protect-kml-kmz-files-from-unauthorized.html
Just some information..
If you pay Google you can have your website run via HTTPS.
If you want something that's a bit more powerful to play with, you might want to try something like GeoServer with an OpenLayers frontend. It can serve your data dynamically from a postgres database as kml (or kmz, gif, jpeg, png, svg, and even pdf) out of the box. You don't have to code a single line to get that done, and it's free.
And indeed, you could draw stuff yourself by using the Google Maps API, but if your page is HTTPS, your users will get a warning about Google's javascript files.
Furthermore, if your browser sucks (at rendering maps), Google does the actual rendering for you server-side, and for that, the actual data will be sent to Google, and a rendered overlay image will be sent back. Maybe you can disable that behaviour, but at least you should be aware of that if you're concerned about security.
As KML files are loaded by Google servers, you can restrict access to KML file only for Google servers.