如何去除二值化眼底血管骨架图像的外围伪影？

我尝试通过提取血管并标记中心像素来定位眼底图像的视盘，现有代码如下：

# Read the image and resize it to 800x800 pixels to reduce the computation time and memory usage
img = load_image(file)
blur = cv2.bilateralFilter(img, 9, 75, 75)
median = cv2.medianBlur(blur, 5)
# Extract the vessels from the image using the sato filter and normalize the image
vessels = sato(median, sigmas=range(1, 10, 2), black_ridges=False)
vessels = (vessels - np.min(vessels)) / (np.max(vessels) - np.min(vessels))
vessels = vessels * 255
vessels = vessels.astype(np.uint8)
_, binary = cv2.threshold(vessels, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
skeleton = morphology.skeletonize(binary)
g, nodes = graph.pixel_graph(skeleton, connectivity=2)
px, distances = graph.central_pixel(g, nodes=nodes, shape=skeleton.shape, partition_size=100)

目前存在的问题：原始图像经处理后，骨架化图像的外围出现大量伪影（对应标注的蓝色区域），完全干扰了血管中心像素的计算，需要去除该区域的伪影。

解决思路与修改后的代码 #

核心思路是先提取眼底的有效圆形区域，屏蔽外围伪影区域，再进行后续的血管提取操作。具体步骤如下：

1. 生成眼底区域掩码：通过灰度化、阈值分割和形态学操作，定位眼底的有效范围
2. 应用掩码：将预处理后的图像与掩码结合，仅保留有效区域内的内容

修改后的代码：

import cv2
import numpy as np
from skimage.filters import sato
from skimage.morphology import skeletonize
from skimage.graph import pixel_graph, central_pixel

# Read the image and resize it to 800x800 pixels
img = load_image(file)
img_resized = cv2.resize(img, (800, 800))

# -------------------------- 新增：提取眼底有效区域掩码 --------------------------
# 转灰度图
gray = cv2.cvtColor(img_resized, cv2.COLOR_BGR2GRAY)
# 大津法分割，得到初始掩码
_, mask = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)
# 形态学闭操作填充内部小孔
kernel_close = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (20, 20))
mask_closed = cv2.morphologyEx(mask, cv2.MORPH_CLOSE, kernel_close)
# 形态学开操作去除外围小噪点
kernel_open = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (10, 10))
mask_clean = cv2.morphologyEx(mask_closed, cv2.MORPH_OPEN, kernel_open)
# 确保掩码是单通道布尔类型
mask_clean = mask_clean.astype(bool)
# -----------------------------------------------------------------------------

# 预处理图像（仅保留有效区域）
blur = cv2.bilateralFilter(img_resized, 9, 75, 75)
# 将外围伪影区域置为背景（黑色）
blur_masked = blur.copy()
blur_masked[~mask_clean] = 0
median = cv2.medianBlur(blur_masked, 5)

# 提取血管并处理
vessels = sato(median, sigmas=range(1, 10, 2), black_ridges=False)
vessels = (vessels - np.min(vessels)) / (np.max(vessels) - np.min(vessels))
vessels = vessels * 255
vessels = vessels.astype(np.uint8)
# 应用掩码，去除外围伪影的血管检测结果
vessels[~mask_clean] = 0

_, binary = cv2.threshold(vessels, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
skeleton = skeletonize(binary)
g, nodes = pixel_graph(skeleton, connectivity=2)
px, distances = central_pixel(g, nodes=nodes, shape=skeleton.shape, partition_size=100)

关键说明 #

• 掩码生成环节采用形态学椭圆核，适配眼底图像的圆形特征，能有效区分眼底有效区域和外围伪影
• 在预处理和血管提取后都应用掩码，确保伪影区域不会参与后续的骨架化和中心像素计算

内容的提问来源于stack exchange，提问作者hascet33