site-logo

JAWS PANKRATION 2024

site-logo
HomeNewsTimetableCfPCommitteePromotionFollow UpPrivacy Policy

Building a Serverless Traffic Monitoring System

Lv200

Lv200

8/24/2024 16:00 (UTC)
Session Info

This describes an event-driven architecture that manages traffic monitoring using AWS Cloud services.

This is an example on using AWS for social change, using this project as an example of how you can solve local challenges using the AWS services

Oluwasegun Adeniyi Adedigba

Oluwasegun Adeniyi Adedigba

- AWS Community Builders -

Session Category
Developer tools
IoT
Storage
AWS Services
Session Materials
    Session Archive
    Session Summary (by Amazon Bedrock)
      The speaker discusses using AWS to create a solution for automatic vehicle recognition at premises entrances. The system replaces manual recording of vehicle details by security personnel. Key components of the solution: 1. ESP32 camera captures images of entering vehicles 2. Images are sent to an S3 bucket via API Gateway 3. Lambda function triggers on new image uploads 4. AWS Rekognition model identifies car brands and detects text (license plates) 5. Information is stored in DynamoDB 6. DynamoDB stream enables notifications (SMS) when vehicle information is processed The speaker trained the recognition model using over 3,000 images of popular car brands in Nigeria. The model achieved high accuracy metrics, including an F1 score of 0.94 and 96% precision. A brief demonstration shows: - S3 bucket containing the dataset used for model training - Uploading new images to trigger the recognition process - DynamoDB table updating with vehicle information The speaker emphasizes the importance of identifying problems in one's environment and using technology to solve them, encouraging listeners to make positive changes in their surroundings.

    Back

    ©JAWS-UG (AWS User Group - Japan). All rights reserved.