Image Processing API

Changelog

• Added crop functionality in Transfrom Image Page.

March 13th 2026

• Finally, Implemented the crop feature in this app.
  It took way longer than I expected. I used react-easy-crop
  for this.

Changelog

• Added Tranform Image page, created new endpoint in backend to get image by id

March 13th 2026

Transform Image Page Finished.

• When you click on a image you get redirected to its transform image where you can test out transformations and get the link to view the image.

12th March 2026
I have created the upload page for my Image Transformer.

• The upload page uploads the image to the application.
• I have added cancel and upload buttons.
• The cancel buttons actually cancels the request using abortController.

What’s been built:

• Implemented JWT Access Tokens[Only in application state.] and JWT Refresh Tokens[securely stored in a http-only cookie] which is the golden standard

• Implemented Routing with react-router.

• Implemented a frontend Login, Register and Images pages.

• Added a PersistLogin Component to get fresh accessTokens to maintain login state.

• Learnt about axios interceptors to automatically retry when accessToken is expired using a new accessToken

Learnt a lot about JWT, next step is fully finish the frontend, implement all backend endpoints.

Changelog

• Added caching, background image removal and GET URL-based Image Transformations.

My thoughts on this devlog :

1. This was a long time coming I finally added caching which reduces the time for already produced images.

2. My friend asked for background removal which is now there too using python [rembg]. I am shocked that there is no background removal in Node.js.

3. Now, GET based transformations to easily embed images anywhere you want.

4. It automatically redirects to AWS S3 bucket image.

14th February 2026
Implement linear version tracking for transformation.
When you perform a transformation on an image, you already uploaded. It stacks.
Hopefully we can implement undo redo functionality with in this in the UI.

Next, caching in AWS S3 already run transformations on the same image to reduce CPU load.

13th February 2026

• Implemented a foreign key between user and image in prisma.
• Implement a robust image transformation endpoint.
• Added view images endpoint.

TODO:

• Add history for every image, so that the user can revert his transformations, or see their history.
• Consider caching to get images quicker.
• Use rabbitMQ for asynchronus requests.

• Implemented JWT Authentication for user registration and user login.
• Added a log out route.
• Added a refresh route to get a new access token.
• Only allowed logged in users to upload images to AWS S3.

I have implemented cloud-based image storage using AWS S3 buckets. When a user uploads his image to the application it will be stored in an S3 bucket.

Used Prisma ORM for the first time. Read the documentation and was able to build a basic login and register system. There is no authentication yet but it will be built.