Artificial intelligence has transformed our digital interactions, and ChatGPT by OpenAI stands at the forefront of this revolution, making image editing a breeze.
In this article, we spotlight three key tools: the Code Interpreter, the Image Editor Plugin, and the MixerBox PhotoMagic Plugin.
The Code Interpreter excels at large-scale image editing, boasting features from watermarking to GIF conversion.
For quick, straightforward edits to online or server-stored images, the Image Editor Plugin shines, offering simple commands for resizing, cropping, and more.
Lastly, the PhotoMagic Plugin enhances overall image quality, specializing in detail recovery and noise reduction.
These three tools constitute a comprehensive image editing suite, each with distinct strengths. Regardless of your editing needs, ChatGPT is here to simplify the process.
Read on as we explore the unique abilities of each tool.
ChatGPT Code Interpreter
The ChatGPT Code Interpreter is well-known for its powerful data analysis capabilities. However, it can also perform batch image editing with incredible features such as:
- face recognition
- image noise reduction
- image to GIF animation
- removing the white background around the subject
These amazing features can be used in batch, as I have explained in my previous article:
Now, let’s discover how to activate this feature and perform some basic image editing.
To switch on the Code Interpreter, you need to have a ChatGPT Plus account. Log in, then click on the three-dot button next to your username at the bottom left to open the ChatGPT settings.
Click “Beta features” on the left, then turn on the Code Interpreter at the bottom right.
Close the window, move your cursor to the “GPT-4” button at the top of the main interface, and select “Code interpreter” from the menu.
You’ve now activated the Code Interpreter.
Starting image processing is as simple as clicking the plus button on the left of the ChatGPT dialog to open your local file browser.
Select the images you’d like to edit or compress them into a zip file for batch processing. I’ll be using the following three portraits as examples.
Batch Image Scaling
Batch scaling images involves resizing multiple images simultaneously, saving time and effort for large image sets. This is useful in scenarios like website optimization, where smaller image files improve loading times. It maintains visual content and aspect ratio across all images.
For this demo, I am currently using the three pictures of the beautiful women mentioned earlier. However, you have the option to include a couple of thousand pictures in a folder and compress them. Just ensure that the total file size does not exceed 512 megabytes.
Let’s scale these three portraits at the same time, enlarging them by a factor of three while maintaining image quality. First, check the original image size.
All three images are currently identical in size.
Start the process by typing the following prompt.
Enlarge the images in the zip package by threefold, and provide a download link for the processed images.
In a short while, ChatGPT will process and compress the images into a downloadable package.
After unzipping, I noticed that the size of all three images had gotten bigger.
And all images were enlarged to three times their original size.
You can also use different Prompts for different scaling effects.
- Scale all images in the zip package to have a width of 800 pixels while maintaining their original aspect ratios.
- Please scale down every image in the given zip file by 50% of their original sizes.
- I want you to enlarge each image within the zip file by a factor of 2.5.
- Can you resize all pictures in this zip package to a height of 600 pixels and preserve their original aspect ratios?
- I need every image in this zip package to be scaled to a resolution of 1920×1080 pixels.
Batch Image Cropping
Batch cropping is a technique that lets you extract specific regions from multiple images at once. It’s particularly useful for processing e-commerce product images to give them a unified look, saving you time while ensuring consistency.
To define the area to crop, use the following prompt:
- Crop each image in the zip package starting from the point (100,100) with a width of 500 pixels and a height of 300 pixels.
- Please extract the region of interest defined by the coordinates (300,200) to (800,600) from every image in the given zip file.
- Can you crop out a square of side 400 pixels from the center of each image in the zip file?
- Crop all photos in the zip package to focus on the area defined by the rectangle starting from the top left corner (0,0) to the bottom right corner (600,400).
- I want to isolate the subject in the middle of every image in the zip package, crop them with a radius of 200 pixels around their centers.
For a fun example, I entered this prompt.
Crop faces from each image in the zip file, compile and compress them, then provide a download link.
ChatGPT efficiently cropped the faces and sensibly renamed the images.
The end result looks like this.
Batch Rotate Images
Batch rotation involves turning a set of images by a certain angle around a specific point. This technique is helpful when you need to uniformly orient multiple photos.
Here’s an example. Enter the following prompt.
Rotate the images in the package -15, 0, and 15 degrees respectively, frame the three rotated images like Polaroids, and arrange into a collage. Provide a download link for the result.
The final result, produced by ChatGPT, looks like this.
Keep in mind that the above example is mainly to demonstrate the potential of the Code Interpreter. Now, let’s look at some practical prompts:
- Rotate each image in the zip package 90 degrees clockwise.
- Please rotate every image in the given zip file 45 degrees counterclockwise.
- Can you rotate each picture in the zip file by 180 degrees around their centers?
- Rotate all images in the zip package around the point (300,300) by an angle of 30 degrees clockwise.
- I need every image in this zip file to be rotated 180 degrees along the horizontal axis.
Batch blurring is about applying a blur effect to multiple images simultaneously, softening details and reducing noise. It’s handy for creating a pleasing aesthetic effect or blurring image backgrounds to minimize distractions.
As a demonstration, I’ll blur the faces of the three portraits using the following prompt.
Apply a 30-pixel Gaussian blur to the faces in the zip package’s portraits, frame them like Polaroids, create a collage, and provide a download link.
The final result will show the images with blurred faces.
There are also other basic but handy prompts for various image editing tasks.
- Apply a Gaussian blur with a radius of 5 pixels to every image in the zip package.
- Blur all images in the zip package using a median filter with a 3×3 kernel.
- Can you apply a blurring effect to each image in the zip file with a sigma of 2.0 using the Gaussian method?
- Please apply a motion blur to all images in the zip file in the horizontal direction with a kernel size of 20.
- I’d like to soften the details in all photos in the zip package, could you apply a blur filter to each of them?
ChatGPT Plugin Image Editor
In some cases, images may be stored online or on a server, and ChatGPT’s code interpreter cannot directly handle online images. In such situations, you can use ChatGPT’s plugin Image Editor, which offers the following functions:
- Upload Pictures: You need to upload your pictures to the cloud space of Image Editor before editing.
- Resize Photo: Specify the width and height (in pixels) for your desired image size.
- Crop Photo: Define the width and height (in pixels) for the section you want to crop.
- Blur Photo: Set the intensity of the blur, from 0 to 2000, with the default set at 100.
- Rotate Photo: Turn your image clockwise, setting the rotation angle from 0 to 360 degrees.
To use this plugin, make sure the ChatGPT plugin features are activated.
Go to the ChatGPT plugin store.
And search for the “Image Editor” plugin to install it.
Once installed, upload images using the following command:
Upload the images for editing:
Don’t forget to replace “[Image URL]” with your image’s actual URL.
The image upload is successful when the Image Editor plugin generates IDs. These let you specify any image during the editing process. I submitted this simple prompt to blur the image:
Blur the photo_1488953 with a strength of 500.
ChatGPT processed the request:
The process is done. Here is your blurred image:
Open blurred image in new tab
This image has been blurred with a strength of 500. You can continue to edit this image or others as needed. The photo ID for this image is “photo_1488953”.
To blur all the images in one go, enter the following prompt, adjusting the intensity as needed.
Blur all the uploaded images with a strength of 500.
MixerBox PhotoMagic ChatGPT Plugin
The ChatGPT Plugin, MixerBox PhotoMagic, enhances image quality in several ways: it recovers details from scanned documents, reduces image noise, and improves image resolution. It’s particularly adept at restoring details in the skin or hair of humans and animals.
Just like the previously mentioned Image Editor plugin, you only need to supply the image URL for PhotoMagic to work its magic. Let’s see its power in recovering details from scanned documents with real-world examples.
Here’s an image showing the comparison of a document before and after processing:
To restore details in scanned images, just use this simple Prompt:
Restore the details of this scanned file: [image url]
Don’t forget to replace “[Image URL]” with your image’s actual URL.
Moreover, PhotoMagic can enhance the resolution of images. If you have a small image that looks blurry when enlarged on a webpage, this plugin can help enhance its resolution. Here’s an example that shows the difference in an image before and after resolution enhancement:
To boost the resolution of an image, use this straightforward prompt:
Enhance the resolution of this photo for me: [image url]
After processing, the image size jumped from 1280×772 to 2560×1544.
In addition to enhancing resolution, PhotoMagic can also reduce image noise. Many images suffer from uneven brightness due to poor lighting during photography, which degrades image quality.
This image below illustrates a comparison of an image before and after noise reduction using PhotoMagic. Notice how the dancer’s rough skin appears smoother after processing:
To decrease image noise, simply feed this prompt to ChatGPT:
Remove the noise from this photo: [image url]
Again, remember to replace “[Image URL]” with your image’s actual URL.
As we’ve explored, ChatGPT and its plugins make image editing accessible to everyone. No longer is this a field for experts only.
The future is here, and it brings the power of artificial intelligence to your fingertips, ready to unleash your creativity in new, exciting ways. So, don’t hesitate to dive in and experiment.
Embrace these tools, redefine your limits, and see what amazing creations you can bring to life.
After all, image editing has become as simple as having a conversation.