| AWS Textra/Rekog |
|---|---|
Author: Irene Cho This plugin uses the official API released from AWS. AWS Textra/Rekog plugin is consisted of two APIs which are Textract and Rekognition – Text Detection. It quickly and easily extracted texts from images/ docs and returns data in the forms of JSON and image file. Prerequisite Users require signing up with AWS service account (AWS key id and secret key). |
Textract
- Extract printed text with optical character recognition (OCR)
- Detect text image (jpeg/png) size up to 10 MB and PDF file size is 500 MB
- Detect up to 3000 words
- Documents can be rotated a maximum of +/- 10% from the vertical axis
- Support only English
- Doesn’t support handwriting docs
Rekognition – Text Detection
- Visual search
- Detect text in .jpeg or .png format images
- Detect up to 50 words
- Text must be within +/- 90 degrees orientation of the horizontal axis
Content
Instructions to create the AWS service account.
1. Set up an Amazon account and create an IAM console:
https://console.aws.amazon.com/iam/
Text from Image
• After signing up the amazon account, sign in ‘IAM user’.
This process provides credentials which can access resources of services.
Text from Image
Text from Image
1) Select 'Add user'.
2) Set user details.
Text from Image
3) Select 'Create group'.
Text from Image
4) Create group.
Text from Image
5) Select ' Next:Tags'.
6) Select 'Next Review'.
7) Select 'Create user'.
8) Click ‘Close’ when you successfully create the user and group.
4. Create an access key.
1) Select Users.
2) Select the Username.
3) Select Security credentials.
4) Create access key.
Text from Image
5) Download the .csv file.
Text from Image
How to set plugin parameters
Text from Image
Plugin Output
1. Image File 2. Out Img File 3. Result type
(Input image file) (Output image file) (Output text)
Return Code
Code | Meaning |
|---|---|
| 0 | NOT Implemented |
| 1 | NOT Implemented |