Workflow name: Product box cover image generator
[Workflow introduction]
This workflow is a bit complicated. The user must first upload 5 images, which include product images and color ID images corresponding to specially made product images (because there are 3 products and the total number of their external surfaces is 5, the color ID is 5 colors. If the number of product display surfaces increases or decreases, it is necessary to manually adjust the workflow mask part accordingly). In addition, 3 cover images must be carefully selected from many pictures. The workflow is divided into two parts, the first part is described here, and its main function is to generate specified images for product boxes. The second part focuses on replacing the background of the product image, which can be achieved with the help of another dedicated workflow.
Background change workflow online running URL:
https://www.runninghub.cn/post/1845758651062743041
https://www.runninghub.ai/post/1845772448322220034
Built-in parameter settings are complete, welcome to experience, wish you a happy time!

[Use scenario]
In the field of product packaging design, when enterprises or designers need to quickly generate a variety of product packaging box cover images, this image generator can show its prowess. Whether it is a new product launch that requires the production of a variety of different styles of packaging previews, or when a series of product packaging is updated, by adjusting the color ID and cover image to batch generate different versions of packaging design, this tool can be used to efficiently complete it. At the same time, for e-commerce merchants who need to change the product background to adapt to different page styles or promotional activity themes when making product display pictures, the second part of the background change workflow can provide a convenient solution to help them obtain high-quality, demand-compliant product packaging box covers and related display images in a short time, thereby enhancing product image and sales appeal.

[Key nodes] Flux1-redux, Flux1-depth

Model version]
FLUX
Model name: Flux1-depth.safetensors

[LoRA model]
None

[ControlNet application]
AnyLine Lineart
DA depth preprocessor

[K sampler]
Sampling method: euler
Scheduler: simple
CFG: 1
Noise reduction: 1