/ece1779A3_Zappa

Auto Tagging Photo Album

Primary LanguagePython

ece1779A3_Zappa

Auto Tagging Photo Album

This web application was developed AWS serveless and Rekognition API.

Users can upload images and the context of these images is captured and tagged automatically. People profiles can also be added so that the next time a photo of a user is uploaded this person is automatically tagged as well

Application Architecture

The application involved 5 main components:

  1. API Gateway
  2. Lamda Functions
  3. S3 Buckets
  4. DynamoDB
  5. Amazon Rekognition API

Web Requests

  • All HTTP requests are mananged by the API Gateway and Lambda functions
  • Image Gallery displays images stored on S3
  • Queries for pictures context or profile are handled by the database

alt text

Upload Images

  • Images are uploaded directly to S3. The S3 upload event triggers the rekognition_labeling function to perform the following tasks on each new image:
  • Call rekognition DetectLabels API to detect labels in S3 object
  • Upload Image Tags to database
  • Check for Faces
  • If a face is found, call recognition Compare_faces
  • Add Face Detected Attribute to Image Table
  • Update tags table with the profile name

alt text

Application Usage

Logging In

alt text

Image Gallery

alt text

Context Filtering

alt text

alt text