Edit in GitHubLog an issue

Product Crop

The Product Crop endpoint facilitates smart cropping for images, automatically detecting the subject and ensuring it remains the focal point of the cropped image. You can identify the product and specify the desired padding for your cropped image.

The /productCrop endpoint can take an input file and apply a crop to it. It doesn't support multilayered PSD files.

Known limitations

There is a known issue that when a figure or portrait is present in the design on a salient object (like a t-shirt, collectible or artwork), the model will perform a crop with the person as the focal area rather than the object itself.

Implementation example

This example shows how to apply the crop with the required padding to an input file:

Copied to your clipboard
curl -X POST \
https://image.adobe.io/pie/psdService/productCrop \
-H "Authorization: Bearer $token" \
-H "x-api-key: $apiKey" \
-H "Content-Type: application/json" \
-d '{
"inputs": [
{
"href": "<SIGNED_GET_URL>",
"storage": "<your_storage>"
}
],
"options": {
"padding": {
"unit": "Pixels",
"width": 10,
"height": 10
}
},
"outputs": [
{
"storage": "<your_storage>",
"type": "image/jpeg",
"href": "<SIGNED_POST_URL>"
}
]
}'
  • Privacy
  • Terms of Use
  • Do not sell or share my personal information
  • AdChoices
Copyright © 2025 Adobe. All rights reserved.