60 lines
1.6 KiB
Markdown
60 lines
1.6 KiB
Markdown
|
|
# Documentation
|
|
- Class name: SaltMaskHistogramEqualizationRegion
|
|
- Category: SALT/Masking/Filter
|
|
- Output node: False
|
|
|
|
该节点对一组遮罩集合中的每个遮罩应用直方图均衡化，以增强遮罩内区域的对比度。它旨在通过调整强度分布来改善每个遮罩内特征的可见度和区分度。
|
|
|
|
# Input types
|
|
## Required
|
|
- masks
|
|
- 需要处理的遮罩集合。每个遮罩都会单独进行增强，通过直方图均衡化来改善其对比度。
|
|
- Comfy dtype: MASK
|
|
- Python dtype: torch.Tensor
|
|
|
|
# Output types
|
|
- MASKS
|
|
- 经过对比度增强的遮罩，这是对每个原始遮罩应用直方图均衡化后的结果。
|
|
- Comfy dtype: MASK
|
|
- Python dtype: torch.Tensor
|
|
|
|
|
|
## Usage tips
|
|
- Infra type: `GPU`
|
|
- Common nodes: unknown
|
|
|
|
|
|
## Source code
|
|
```python
|
|
class SaltMaskHistogramEqualizationRegion:
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"masks": ("MASK",),
|
|
}
|
|
}
|
|
|
|
CATEGORY = f"{NAME}/Masking/Filter"
|
|
|
|
RETURN_TYPES = ("MASK",)
|
|
RETURN_NAMES = ("MASKS",)
|
|
|
|
FUNCTION = "histogram_equalization"
|
|
|
|
def histogram_equalization(self, masks):
|
|
regions = []
|
|
for mask in masks:
|
|
pil_image = mask2pil(mask.unsqueeze(0))
|
|
image_array = np.array(pil_image.convert('L'))
|
|
equalized = cv2.equalizeHist(image_array)
|
|
equalized_pil = ImageOps.invert(Image.fromarray(equalized))
|
|
region_tensor = pil2mask(equalized_pil)
|
|
regions.append(region_tensor)
|
|
|
|
regions_tensor = torch.cat(regions, dim=0)
|
|
return (regions_tensor,)
|
|
|
|
```
|