51 lines
1.6 KiB
Markdown
51 lines
1.6 KiB
Markdown
# Documentation
|
|
- Class name: SvgImage
|
|
- Category: ♾️Mixlab/Image
|
|
- Output node: False
|
|
- Repo Ref: https://github.com/shadowcz007/comfyui-mixlab-nodes.git
|
|
|
|
该节点旨在处理SVG图像，将其转换为适合在神经网络框架中进一步操作和分析的张量格式。它强调SVG数据的转换和适配，以便与深度学习模型集成。
|
|
|
|
# Input types
|
|
## Required
|
|
- upload
|
|
- ‘upload’参数对节点的运行至关重要，它包含了SVG图像数据和相关的元数据。它直接影响节点处理和转换SVG为张量的能力。
|
|
- Comfy dtype: Dict[str, Any]
|
|
- Python dtype: Dict[str, Any]
|
|
|
|
# Output types
|
|
- IMAGE
|
|
- ‘IMAGE’输出提供了输入SVG图像的张量表示，可用于深度学习环境中的进一步处理。
|
|
- Comfy dtype: Tensor
|
|
- Python dtype: torch.Tensor
|
|
- layers
|
|
- ‘layers’输出包含了从SVG数据中提取的单个组件或图层，可用于对图像结构进行详细分析或操作。
|
|
- Comfy dtype: List[Any]
|
|
- Python dtype: List[Any]
|
|
|
|
# Usage tips
|
|
- Infra type: CPU
|
|
|
|
# Source code
|
|
```
|
|
class SvgImage:
|
|
|
|
@classmethod
|
|
def INPUT_TYPES(s):
|
|
return {'required': {'upload': ('SVG',)}}
|
|
RETURN_TYPES = ('IMAGE', 'LAYER')
|
|
RETURN_NAMES = ('IMAGE', 'layers')
|
|
FUNCTION = 'run'
|
|
CATEGORY = '♾️Mixlab/Image'
|
|
INPUT_IS_LIST = False
|
|
OUTPUT_IS_LIST = (False, True)
|
|
|
|
def run(self, upload):
|
|
layers = []
|
|
image = base64_to_image(upload['image'])
|
|
image = image.convert('RGB')
|
|
image = pil2tensor(image)
|
|
for layer in upload['data']:
|
|
layers.append(layer)
|
|
return (image, layers)
|
|
``` |