import numpy as np from PIL import Image import torch import cv2 def numpy2tensor(image: Image.Image): return torch.from_numpy(image).unsqueeze(0) class LightColorCulling: @classmethod def INPUT_TYPES(self): return { "required": { "image": ("IMAGE",), "white_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 0.9, "step": 0.01}), "transition_range": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0, "step": 0.01}), }, } RETURN_TYPES = ("IMAGE", ) RETURN_NAMES = ("image", ) FUNCTION = 'Preprocessing' CATEGORY = 'StickerEdit/LightColorCulling' def Preprocessing(self, image, white_threshold=0.7, transition_range=0.2): image_np = (image.cpu().numpy().squeeze()*255).astype(np.uint8) # 检查图像是否有透明通道 has_alpha = image_np.shape[2] == 4 # 转换为灰度图 gray = cv2.cvtColor(image_np, cv2.COLOR_BGR2GRAY) # 归一化灰度图像到0-1范围 normalized_gray = gray / 255.0 # 创建结果图像 result_image = image_np.copy() # 创建透明度通道 if has_alpha: b_channel, g_channel, r_channel, alpha_channel = cv2.split(result_image) else: alpha_channel = np.ones_like(gray, dtype=float) * 255 b_channel, g_channel, r_channel = cv2.split(result_image) # 计算渐变透明度 mask = normalized_gray >= white_threshold transition_mask = (normalized_gray >= (white_threshold - transition_range)) & (normalized_gray < white_threshold) alpha_channel[mask] = 0 alpha_channel[transition_mask] = (1 - (normalized_gray[transition_mask] - (white_threshold - transition_range)) / transition_range) * 255 # 更新结果图像的alpha通道 # import pdb;pdb.set_trace() result_image = cv2.merge((b_channel, g_channel, r_channel, alpha_channel.astype(np.uint8))) res_tensor = numpy2tensor(result_image/255.0) return (res_tensor,) NODE_CLASS_MAPPINGS = { "LightColorCulling": LightColorCulling, } NODE_DISPLAY_NAME_MAPPINGS = { "LightColorCulling": "LightColorCulling", }