PhotopilotPhotopilot
Dashboard
View demo
Consultation
EN
NL Nederlands EN English DE Deutsch FR Français
All cases Customer story

Duifhuizen

Bags and suitcases believably to scale in every AI image

At Duifhuizen, the range spans from small leather goods to large travel suitcases. Here, scale determines if an AI image feels believable: a key pouch shouldn't look like a handbag, and a cabin suitcase must remain visibly more compact than a 75-centimeter spinner.

Photopilot combines the source packshot with width, depth, and height from the article data. For each product, the visible object area is re-measured and placement is dynamically calculated. This creates a warm, recognizable image line without building a separate Photoshop template for each size.

The production review of September 2026 also shows that images and content are used side by side: over 500 AI images for more than 300 items, and more than 800 items with generated content, most of which were accepted after review.

www.duifhuizen.nl
White travel suitcase believably to scale in a bright hotel room
How it works in practice Duifhuizen
Modules

This is how this case connects to the rest of Photopilot.

Dynamic scalingObject detection, actual product dimensions, and fixed margins in one repeatable image standard.View moduleAI campaignsFrom product packshot to mood image or model image within the same article workflow.View moduleProduct contentIn production, 585 AI content edits have been recorded. The output includes product names, descriptions, meta descriptions, and search terms.View module
Real AI images from production

From product size to believable scene

Suitcases, weekend bags, and small leather goods shouldn't all appear the same size. Photopilot uses the real article dimensions to dynamically scale each product and place it in one recognizable visual language.

White Enrico Benetti suitcase of 75 centimeters in a bright hotel room with visible dimensions
AI scene from Duifhuizen production 46 × 27 × 75 cm
Source packshotSource packshot of a white Enrico Benetti suitcase
How dynamic scaling works

Four steps, without starting over for each image

  1. 01

    Article dimensions as anchor

    Width, depth, and height are included from the Duifhuizen article data. This ensures each image starts with a known size.

  2. 02

    Recognize product area

    Photopilot detects the product in the source image, removes empty white space, and determines the base point and visible product box.

  3. 03

    Calculate scale per product

    Placement is recalculated for this size. A compact bag doesn't get the same zoom factor as a 75-centimeter suitcase.

  4. 04

    Result next to the source

    Source and AI result remain visible with the same article. The team can compare variants and only proceed with the best image.

Recently out of production

New suitcases and backpacks, directly from the Duifhuizen environment

These recent results are selected from the Duifhuizen production environment. This case not only demonstrates the scaling principle but also shows what is actually being generated.

Recent AI image of a white Enrico Benetti suitcase to scale in a light interior
AI scene

Enrico Benetti Spinner 75

Recent AI model image with a beige Castelijn & Beerens laptop backpack
AI model image

Castelijn & Beerens laptop backpack

Recent AI model image of a woman with a black Michael Kors handbag at true scale
AI model image

Michael Kors handbag

Reviewed production data

Images and content, backed by real usage data

This snapshot comes from the Duifhuizen production environment and was reviewed on September 18, 2026. The figures show the actual assortment and recorded usage up to that point.

14.897
active items in the Duifhuizen production environment
13.327
items of the bag type alongside accessories, belts, and other product groups
500+
AI images for 300+ items model images and scenes
800+
items with content output most accepted after review
750+ AI content edits recorded

The production data shows content output for product name, description, meta description, and search terms. This way, the same item data can feed both the images and the sales content.

  • 547 items with description and meta description
  • 513 items with generated product name
  • Human review remains part of the workflow
One image line, different formats

From 16 to 75 centimeters

The images below come from the same production environment. The source packshot remains visible, allowing you to directly verify product shape, color, and scale.

Beige Enrico Benetti weekend bag in a bright hotel room, generated from a Duifhuizen packshot
Source packshot · Weekend bag
Enrico Benetti

Weekend bag

55 × 25 × 25 cm
Black Enrico Benetti backpack believably to scale in a warm interior
Source packshot · Compact backpack
Enrico Benetti

Compact backpack

22 × 11 × 27 cm
Black Hide & Stitches shoulder bag believably to scale in a light interior
Source packshot · Shoulder bag
Hide & Stitches

Shoulder bag

23 × 4 × 16 cm
Also shown with a person

A suitcase stays a 76-centimetre suitcase

The same size data also helps with AI model images. This gives the customer not only ambiance but also a useful sense of the actual product size.

  • Product shape and color remain recognizable
  • Person and product are in a believable proportion
  • Result remains linked to the source article
Woman of 175 cm next to a dark blue 76 cm Suitsuit suitcase
AI model image
Source packshot of a dark blue Suitsuit suitcase
Source packshot

Real product dimensions as scale anchor

Width, depth, and height remain linked to the article. These dimensions guide the proportion between product, person, interior, and open space.

View module

AI scenes from an existing packshot

A standalone product image becomes the source for a context image. Product shape, color, and recognizable details remain visible for verification.

View module

Evaluate variants next to the source

Source image and AI results stay with the same article. The team decides what can proceed and keeps exceptions out of the publication flow.

View this step
More customer stories

See how other fashion teams organize their workflow.

FAQ

Frequently Asked Questions

What problem does Photopilot solve for Duifhuizen?

Duifhuizen's assortment ranges from small leather goods to large travel cases, and scale determines whether an AI image is credible. Photopilot uses the actual width, depth, and height from the item data, ensuring a shoulder bag, weekend bag, and 75 cm suitcase remain recognizably proportionate.

AI campaigns

What does Duifhuizen need to create AI images at scale?

The Duifhuizen case in Photopilot starts with two things: a freestanding source packshot and product dimensions linked to the article. This eliminates the need for a separate Photoshop template per product size, and the same image standard works for both a large suitcase and a small bag.

Packshot editing

Does the Duifhuizen team still review the AI images themselves?

Yes. At Duifhuizen, source images and AI results remain with the same item in Photopilot, allowing for verification of product shape, color, and scale. The team decides what can proceed and keeps exceptions out of the publication flow; human oversight remains part of the workflow.

How it works

What figures are in the Duifhuizen case?

The snapshot from the Duifhuizen production environment was reviewed on September 18, 2026: over 500 AI images for more than 300 items and more than 800 items with content output, most of which were accepted after review. The content output includes product names, descriptions, meta descriptions, and search terms.

Content