Business Case - Users do not want to type info in a form.
Upload a document image and expect the fields to filled automatically e.g. upload license, ID etc. They are ready to modify if the auto tool does not capture the info well. In short the customer needs OCR and do not want to store their info in a database to be read manually.
Solution : Google Cloud vision API provides the same, and below is an explanation of how to do it quickly and a working example below.
Try to upload an image with text, and the output will extract a text from image.
NOTE: if you are using mobile, make sure the image is good quality and clear (not like whatsapp images) , else you might get cloud vision error. If possible, use desktop to test the below demo.
Prerequisite :
How to do ?
Step 1 : Initiate the vision as shown below
Step 2 : Stream the file to vision and get the HTML response. No need to store the image, unless required. Working code as below.
That's it, its simple.
Upload a document image and expect the fields to filled automatically e.g. upload license, ID etc. They are ready to modify if the auto tool does not capture the info well. In short the customer needs OCR and do not want to store their info in a database to be read manually.
Solution : Google Cloud vision API provides the same, and below is an explanation of how to do it quickly and a working example below.
Try to upload an image with text, and the output will extract a text from image.
NOTE: if you are using mobile, make sure the image is good quality and clear (not like whatsapp images) , else you might get cloud vision error. If possible, use desktop to test the below demo.
Prerequisite :
- Node JS hands on and Node JS app on Heroku (AWS gives some issues sometimes with the GCloud services, so I used Heroku for the POC)
- Google project is created with the google key file for using cloud vision service
- You can check this link if not clear about the pre-reqisites https://cloud.google.com/functions/docs/tutorials/ocr
How to do ?
Step 1 : Initiate the vision as shown below
Step 2 : Stream the file to vision and get the HTML response. No need to store the image, unless required. Working code as below.
That's it, its simple.
You can certainly add plus launch and many choices out there. People are prompted to customize in case the automatic resource would not take the information Digital Ocean vs AWS perfectly. In brief, the shopper demands OCR and don't prefer to save its information and facts inside of a collection that they are learning yourself.
ReplyDelete