Relight Composited Product
Upload a composited image of your product, draw a mask in the mask editor and relight your product into the scene.
简介
The 'Relight Composited Product' workflow in ComfyUI is designed to enhance product images by allowing users to adjust the lighting of a specific area within a composited image. This is particularly useful for product photographers and e-commerce businesses looking to create visually appealing images that stand out.
The workflow leverages the Qwen-Image-Edit model, known for its advanced image editing capabilities, to achieve realistic lighting adjustments. By using the 'comfyui-inpaint-cropandstitch' custom node, users can precisely target areas of the image to relight, ensuring seamless integration with the existing scene.
This workflow is beneficial for improving the visual aesthetics of product images without the need for complex photo editing software.
你需要提供
- 正在读取工作流
在 ComfyUI 中运行。 此工作流无法在本页面运行。你可以在 Comfy Cloud 中打开（已预先加载所需内容），或下载后在自己的 ComfyUI 中使用。
在 Cloud 中试用查看工作流内部
节点图的只读预览。在 Comfy Cloud 中打开或下载后即可编辑节点。
Loading the graph
Open full-size workflow preview- 作者
- Rob
需要自定义节点
- 正在读取工作流
通过代码使用此工作流
尽可能根据工作流图和输入自动生成代码，否则显示已保存的示例。运行代码可能产生费用，请先核对凭据和端点。
What you need
- Runs on
- Comfy Cloud
- 密钥
- COMFY_API_KEY
需要付费的 Cloud 方案和额度。
# Python 3.10+: pip install comfy-sdk==0.4.0
import os
from comfy_sdk import Comfy
api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
"103": {
"class_type": "LoadImage",
"inputs": {
"image": "comfy_cloud_bottle.png"
},
"_meta": {
"title": "Load Image + Draw Square Mask"
}
},
"208": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "ComfyUI",
"images": [
"224:213",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"224:119": {
"class_type": "UNETLoader",
"inputs": {
"unet_name": "qwen_image_edit_2509_fp8_e4m3fn.safetensors",
"weight_dtype": "default"
},
"_meta": {
"title": "Load Diffusion Model"
}
},
"224:131": {
"class_type": "LoraLoaderModelOnly",
"inputs": {
"lora_name": "Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors",
"model": [
"224:89",
0
],
"strength_model": 1
},
"_meta": {
"title": "Load LoRA"
}
},
"224:211": {
"class_type": "TextEncodeQwenImageEditPlus",
"inputs": {
"clip": [
"224:38",
0
],
"image1": [
"224:221",
1
],
"image2": [
"224:221",
1
],
"prompt": "Image blending, correcting the product's perspective angle and light and shadow to integrate the product into the background",
"vae": [
"224:39",
0
]
},
"_meta": {
"title": "Text Encode Qwen Image Edit Plus"
}
},
"224:212": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"224:222",
0
],
"vae": [
"224:39",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"224:213": {
"class_type": "InpaintStitchImproved",
"inputs": {
"inpainted_image": [
"224:212",
0
],
"stitcher": [
"224:221",
0
]
},
"_meta": {
"title": "✂️ Inpaint Stitch (Improved)"
}
},
"224:215": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"224:212",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"224:216": {
"class_type": "TextEncodeQwenImageEditPlus",
"inputs": {
"clip": [
"224:38",
0
],
"image1": [
"224:221",
1
],
"image2": [
"224:221",
1
],
"prompt": "",
"vae": [
"224:39",
0
]
},
"_meta": {
"title": "Text Encode Qwen Image Edit Plus"
}
},
"224:217": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"224:221",
1
],
"vae": [
"224:39",
0
]
},
"_meta": {
"title": "VAE Encode"
}
},
"224:218": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"224:131",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"224:219": {
"class_type": "CFGNorm",
"inputs": {
"model": [
"224:218",
0
],
"pre_cfg": False,
"strength": 1
},
"_meta": {
"title": "CFGNorm"
}
},
"224:220": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"224:221",
1
]
},
"_meta": {
"title": "Preview 1024x1024 Crop Image"
}
},
"224:221": {
"class_type": "InpaintCropImproved",
"inputs": {
"context_from_mask_extend_factor": 1,
"downscale_algorithm": "bilinear",
"extend_down_factor": 1,
"extend_for_outpainting": False,
"extend_left_factor": 1,
"extend_right_factor": 1,
"extend_up_factor": 1,
"image": [
"103",
0
],
"mask": [
"103",
1
],
"mask_blend_pixels": 16,
"mask_expand_pixels": 0,
"mask_fill_holes": True,
"mask_hipass_filter": 0.1,
"mask_invert": False,
"output_padding": "0",
"output_resize_to_target_size": True,
"output_target_height": 1024,
"output_target_width": 1024,
"preresize": False,
"preresize_max_height": 16384,
"preresize_max_width": 16384,
"preresize_min_height": 1024,
"preresize_min_width": 1024,
"preresize_mode": "ensure minimum resolution",
"upscale_algorithm": "bicubic"
},
"_meta": {
"title": "✂️ Inpaint Crop (Improved)"
}
},
"224:222": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"224:217",
0
],
"model": [
"224:219",
0
],
"negative": [
"224:216",
0
],
"positive": [
"224:211",
0
],
"sampler_name": "euler",
"scheduler": "simple",
"seed": 1032597172973225,
"steps": 4
},
"_meta": {
"title": "KSampler"
}
},
"224:223": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"224:213",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"224:38": {
"class_type": "CLIPLoader",
"inputs": {
"clip_name": "qwen_2.5_vl_7b_fp8_scaled.safetensors",
"device": "default",
"type": "qwen_image"
},
"_meta": {
"title": "Load CLIP"
}
},
"224:39": {
"class_type": "VAELoader",
"inputs": {
"vae_name": "qwen_image_vae.safetensors"
},
"_meta": {
"title": "Load VAE"
}
},
"224:89": {
"class_type": "LoraLoaderModelOnly",
"inputs": {
"lora_name": "Qwen-Image-Edit-2509-Fusion.safetensors",
"model": [
"224:119",
0
],
"strength_model": 1
},
"_meta": {
"title": "Load LoRA"
}
}
})
# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)常见问答
While designed for product images, this workflow can be adapted for any composited image where specific relighting is needed.
Ensure that the mask is accurately drawn and make use of the comfyui-inpaint-cropandstitch node to achieve a seamless blend.
Yes, the workflow is user-friendly and does not require advanced photo editing skills, making it accessible for beginners.
The LoadImage node supports common formats like JPEG and PNG, ensuring compatibility with most image files.