电商产品自适应出图设计ComfyUI 工作流节点

简介

自适应尺寸，各种形状或尺寸的产品可以自动适应固定的分辨率。例如，如果产品在上传图片中所占比例较大，则重新计算为“True”后，主体分辨率会自动调整为缩小。
您只需单击按钮即可将输出选择为横向或纵向模式。
有七种内置样式可供选择，您可以通过输入 0 到 6 之间的数字来实现一种样式。依旧可以在原图上进行遮罩编辑来绘制出遮挡区域，且完全不受分辨率或横竖屏影像，依旧自适应。
你可以通过百分比来调控原图或最终生成图像的融合度

工作流节点截图

附件文件一览

模型
Checkpoints
juggernautXL_juggernautX.safetensors👉前往下载

juggernaut_reborn.safetensors👉前往下载

sai_xl_depth_256lora.safetensors👉前往下载

controlnet-tile-sdxl-1.0.safetensors👉前往下载

IC-Light👉前往下载

LoRAs
Majicproduct_P3.safetensors👉前往下载

内置节点
Anything Everywhere
Anything Everywhere3
AreaCalculator
Display Any (rgthree)
GetNode
Image Comparer (rgthree)
LogicUtil_MergeString
MiniCPM_VQA
Note
SetNode
SimpleComparison+
SimpleCondition+
SimpleMathBoolean+

自定义节点
BrushNet
– BrushNetLoader
– BrushNet
ComfyUI
– EmptyImage
– CheckpointLoaderSimple
– ImageToMask
– VAEDecode
– MaskComposite
– MaskToImage
– VAEEncode
– LatentUpscaleBy
– CLIPTextEncode
– ControlNetApplyAdvanced
– ControlNetLoader
– PreviewImage
– GrowMask
– LoadImage
– KSampler
– LoraLoaderModelOnly
ComfyUI Essentials
– ImageResize+
– GetImageSize+
– RemBGSession+
– ImageRemoveBackground+
ComfyUI Impact Pack
– ImpactStringSelector
ComfyUI Layer Style
– LayerUtility: ImageBlendAdvance
– LayerMask: MaskInvert
– LayerUtility: ImageBlend
ComfyUI Nodes for Inference.Core
– Zoe_DepthAnythingPreprocessor
ComfyUI-IC-Light
– LoadAndApplyICLightUnet
– ICLightConditioning
– DetailTransfer
ComfyUI-Logic
– Float
– Int
comfyui-mixlab-nodes
– FeatheredMask
Efficiency Nodes for ComfyUI Version 2.0+
– KSampler (Efficient)
pythongosssss/ComfyUI-Custom-Scripts
– ShowText|pysssss
Use Everywhere (UE Nodes)
– Seed Everywhere