Project Horizon

Imagine a world where images are processed and transformed effortlessly in the cloud.
Project Horizon is an image processing pipeline that optimize images uploaded by users, making them smaller and faster to load without sacrificing visual quality.

Table of Contents

• System Requirements
  • Functional Requirements
  • Non-Functional Requirements
• System Design
• Balance between speed, storage, and image quality
  • Benchmarks
• How to deploy and run Project Horizon
  • Deploy the infrastructure to AWS
  • CLI Usage
  • HTTP API Usage
  • Delete the infrastructure
• How to run unit tests
• CI/CD Pipeline

System Requirements

Functional Requirements

• The service is automatically triggered when an image is uploaded to a source storage. (eg an object storage).
• The service processes the image uploaded to the source storage and uploads it to a destination storage.
• The processed image must be lighter and faster to load.
• The system must ensure good quality of the processed images.

Non-Functional Requirements

• The system ensures efficient image processing to minimize execution time and costs.
• The system includes basic error handling to ensure the robustness of your function.

System Design

The architecture is easily deployable on AWS through CloudFormation. The system leverages S3 and Lambda services which are integrated through S3 events when a resource is uploaded to the source bucket.
It can be used through the AWS CLI by uploading/downloading images directly from the S3 buckets.

Development to enable the service through the HTTP protocol is a work in progress.

CLI interaction example:

HTTP interaction example:

Balance between speed, storage, and image quality.

The system uses the SixLabors.ImageSharp library to manipulate images and is configured to balance storage efficiency and image quality.

The maximum size of the optimized images is 1.000x1.000 pixels, and the Lanczos algorithm is used for downscaling, ensuring good image quality. You can configure it in the appSettings.json file by modifying the MaxImageDimension property.

To prevent memory issues on Lambda, the property MaxImageSizeInBytes in the appSettings.json file has been configured to set the maximum image size (in bytes) for processing. If a larger image is uploaded to the source storage bucket, it will not be processed by the Lambda.

As the conversion algorithm, WebP in Lossless mode has been used. WebP is a good choice because it offers efficient compression without compromising image quality and is supported by all major web browsers.

Benchmarks (Maximum image dimension: 1000 pixels)

Filename	Original Dimension (pixels)	Original Size (MB)	Final Dimension (pixels)	Final Size (MB)	Lambda Billed Duration (ms)
source_800_800.png	800x800	1.3	800x800	0.77	7228
source_1920_1080.png	1920x1080	3.9	1000x562	0.70	7632

How to deploy and run Project Horizon

For deploying and running Project Horizon, you need to have:

• An AWS account
• AWS CLI configured

Deploy the infrastructure to AWS

• From the root of the project run ./deploy.sh (default region: eu-central-1, default env: dev)
• Example for a custom deployment ./deploy.sh us-west-2 prod

CLI Usage

• Upload an image on the source S3 bucket.
  • Run aws s3 cp <local_path>/image_name.png s3://<YOUR_AWS_ACCOUNT_ID>-source-bucket
• Download the optimized image from the destination S3 bucket.
  • Run aws s3 cp s3://<YOUR_AWS_ACCOUNT_ID>-destination-bucket/image_name.webp <local_path>

HTTP API Usage

1. Get a presigned URL to upload an image:

curl -X GET https://<API_GATEWAY_URL>/prod/presigned-url

Response:

{
    "imageId": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
    "uploadUrl": "https://s3.../presigned-url"
}

2. Upload your image using the presigned URL:

curl -X PUT "<UPLOAD_URL_FROM_RESPONSE>" \
-H "Content-Type: image/png" \
--data-binary @your-image.png

3. Get information and presigned URL to download the optimized image:

curl -X GET https://<API_GATEWAY_URL>/prod/optimized-image/{imageId}

Response:

{
    "imageId": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
    "status": "Success",
    "url": "https://s3.../optimized-image-presigned-url",
    "dateTime": "2025-11-26T13:02:32.123Z"
}

4. Download the optimized image:

curl -X GET "<DOWNLOAD_URL_FROM_RESPONSE>" \
-o optimized-image.webp

There are some images provided for you inside the SourceImages folder.

Delete the infrastructure

• From the root of the project run ./delete.sh (default region: eu-central-1, default env: dev)

How to run unit tests

• Run dotnet test

CI/CD Pipeline

This project uses GitHub Actions for Continuous Integration and Continuous Deployment.

• CI: Every push to the main branch and every pull request trigger automated tests, ensuring the code compiles successfully and all tests pass.
• CD: WIP