Removing background and resizing images
AI image generation doesn’t stop once an image is created. To use your image in real-world settings—like websites, social media posts, or print designs—you often need to refine it. That’s where background removal, resizing, and layout adjustment come in.
In this section, we’ll explore simple but essential techniques to make your generated images more flexible and practical. Removing a background lets you place a character or object into different scenes. Resizing ensures the image fits specific formats, like YouTube thumbnails or Instagram posts, without distortion. And by combining these steps, you can create polished visuals ready for any use.
Let’s get started with background removal and build from there.
Removing background
Background removal is one of the simplest and most impactful edits. It isolates your subject and allows you to reuse it in multiple contexts. Many platforms provide this as a one-click tool, while some let you do it through a natural conversation.
One of the easiest ways to remove the background from an image is by using ChatGPT 4o. If you’re working within a conversational interface that supports image understanding (like ChatGPT with the latest image tools enabled), you can simply upload an image and ask it to "remove the background from this image." The model will return the subject with a transparent or white background.
Case example: Furferno
In Chapter 3, we created a whimsical character named Furferno—a glowing, blue-furred creature with fiery eyes and a swirling tail. Let’s say we want to isolate Furferno to use him across different backgrounds.
We upload the image to ChatGPT 4o and ask:
Can you remove the background of this image, keeping only the character?
Within seconds, the image is returned with the background removed. You now have a clean subject—Furferno—ready for reuse across scenes, from sci-fi cities to enchanted forests.
If you don’t have access to ChatGPT 4o, no worries. You can use these free tools:
Both platforms allow you to upload an image and automatically remove its background, no registration required.
Once you've removed the background, you can replace it with something new—whether that’s a landscape, a textured surface, or an abstract gradient. This is especially useful when you want to generate consistent themes for campaigns or digital compositions.
Adjusting image size
Different platforms call for different image dimensions. Instagram favors square (1:1), YouTube thumbnails require 16:9, and vertical reels need a 9:16 format. The challenge is resizing your image without compromising its look.
If you need to change an image’s aspect ratio but keep its aesthetic, ChatGPT 4o can again help.
Case example
Suppose you have a background image—a soft, monochrome gradient—that was originally generated in a vertical format. You want to convert it into a 3:2 ratio for a desktop wallpaper.
Prompt example:
Can you resize this image to 3:2 aspect ratio?
The model resizes the image intelligently, extending edges or applying AI-based padding where necessary, ensuring it maintains the same atmosphere.
Integrating character and background
Let’s now combine the two elements we worked with: the Furferno character (cut out from the original image) and the monochrome 3:2 background. You can ask ChatGPT but its capability to place in a precise location is not ready. As the technology evolves very fast, soon or later, we’ll be able to integrate images with professional precisions without switching platforms. Use an image editing tool, to combine the two images.
Removing backgrounds and resizing images are two of the most fundamental skills you will practice again and again on your AI image generation journey. Whether you’re preparing visuals for a website, designing a product mockup, or simply refining an illustration, mastering these steps will save time and improve the overall quality of your work.