CavinHuang-comfyui-nodes-docs/docs/TilePreprocessor_Provider_for_SEGS __Inspire.md

Documentation

• Class name: TilePreprocessor_Provider_for_SEGS __Inspire
• Category: InspirePack/SEGS/ControlNet
• Output node: False

该节点为SEGS（语义边缘引导合成）提供预处理服务，应用了一种平铺预处理技术。它旨在通过金字塔上采样调整分辨率来增强输入数据，以提高SEGS应用中的性能表现。

Input types

Required

• pyrUp_iters
  • 指定金字塔上采样的迭代次数，影响处理后图像的最终分辨率。较高的迭代次数会产生更精细的分辨率，从而可能提高SEGS输出的质量。
  • Comfy dtype: INT
  • Python dtype: int

Output types

• segs_preprocessor
  • 返回一个经过预处理的对象，准备用于SEGS处理，封装了在平铺预处理阶段所做的调整。
  • Comfy dtype: SEGS_PREPROCESSOR
  • Python dtype: TilePreprocessor_wrapper

Usage tips

• Infra type: CPU
• Common nodes: unknown

Source code

class TilePreprocessor_Provider_for_SEGS:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {'pyrUp_iters': ("INT", {"default": 3, "min": 1, "max": 10, "step": 1})}}
    RETURN_TYPES = ("SEGS_PREPROCESSOR",)
    FUNCTION = "doit"

    CATEGORY = "InspirePack/SEGS/ControlNet"

    def doit(self, pyrUp_iters):
        obj = TilePreprocessor_wrapper(pyrUp_iters)
        return (obj, )