如果单个图像中有多个 ROI，如何计算每个单独 ROI 的 RGB 平均值？

Question

我有一张 RGB 格式的 3 通道图像，其中包含许多建筑物作为感兴趣区域。图像的背景是全黑的（零像素）。我尝试计算每个单独 ROI 的颜色特征，并将它们保存在 CSV 文件中，但我的代码仅将单个 RGB 值作为每个通道的平均值，因为我需要保存 R、G、B 和 RGB 平均值CSV 文件中的每个 ROI。任何指导将不胜感激。

import numpy as np
import pandas as pd
from skimage import io

image = io.imread("/content/drive/MyDrive/277.png")

current_roi = 1
start_pixel = None
roi_mean_values = []

for y in range(image.shape[0]):
for x in range(image.shape[1]):

pixel_value = image[y, x]

     if np.any(pixel_value > 0):  
           if start_pixel is None:
                 start_pixel = (x, y)
           elif start_pixel is not None:
               end_pixel = (x - 1, y)  # End of the current ROI
               roi = image[start_pixel[1]:end_pixel[1] + 1, start_pixel[0]:end_pixel[0] + 1]
    
               mean_rgb = np.mean(roi, axis=(0, 1))
               roi_mean_values.append({'ROI': current_roi,
                                    'Mean_R': mean_rgb[0],
                                    'Mean_G': mean_rgb[1],
                                    'Mean_B': mean_rgb[2]})
    
              current_roi += 1
              start_pixel = None

roi_mean_df = pd.DataFrame(roi_mean_values)

roi_mean_df.to_csv("roi_mean_values.csv", index=False)

print("RGB mean values for individual buildings (ROIs) saved to 'roi_mean_values.csv'.")

Answer 1

据我了解，投资回报率是由输入图像的“连接组件”给出的。每个连接的组件都是一个 ROI。 您可以在不迭代 ROI 的情况下解决这个问题，如果有很多 ROI，这可能会变得相当昂贵。不过，这段代码确实显式地迭代了像素，这在 Python 中很慢。也许，迭代 ROI 并使用编译函数处理所有像素会产生更快的 Python 代码，尽管实际上正在完成更多的工作。但是用编译语言编写的该算法是您可以编写的最有效的算法。

import numpy as np
import skimage

image = skimage.io.imread("/Users/cris/Downloads/G4IXb.png")
n_channels = image.shape[2]

# Find binary image separating background and objects
mask = np.any(image > 0, axis=2)

# Label this mask image, each ROI gets a unique label
lab, n_labs = skimage.measure.label(mask, return_num=True)

# Create arrays where we'll collect the information
sum = np.zeros((n_labs + 1, n_channels))
number = np.zeros(n_labs + 1)

# Loop over all pixels.
for ii in range(image.shape[0]):
    for jj in range(image.shape[1]):
        # For each pixel, look at the label, and add information to
        # the arrays for that label.
        index = lab[ii, jj]
        sum[index, :] += image[ii, jj, :]
        number[index] += 1

# We're done. The mean color is the sum divided by the number of pixels.
mean_colors = sum / number[:, None]

让我们看看平均颜色是什么样的：

# Paint an image with the mean colors
mean_colors = np.round(mean_colors).astype(np.uint8)
out = np.empty_like(image)
for ii in range(image.shape[0]):
    for jj in range(image.shape[1]):
        out[ii, jj] = mean_colors[lab[ii, jj]]

import matplotlib.pyplot as plt
plt.subplot(1, 2, 1)
plt.imshow(image)
plt.subplot(1, 2, 2)
plt.imshow(out)
plt.show()