提供能实现基于概率分布的超声图像与自然图像性质差异分析的完整python代码，超声图像的文件路径是D:\zzz\na5，自然图像的文件路径是D:\zzz\us5,两种图像都有778张，请注意每张图像的长和宽不都一样可能需要调整，需要对比的特征有形态和纹理，需要统计所有结果进行差异性检验和分析，需要可视化使结果尽量明显

以下是基于概率分布的超声图像与自然图像性质差异分析的完整Python代码。请注意，本代码需要安装scikit-image、numpy、matplotlib和scipy库。

import os
import numpy as np
from skimage import io, feature
from scipy.stats import kstest
import matplotlib.pyplot as plt


def load_images(folder):
    images = []
    for filename in os.listdir(folder):
        img = io.imread(os.path.join(folder, filename))
        if img is not None:
            images.append(img)
    return images


def preprocess_images(images, target_size):
    preprocessed_images = []
    for img in images:
        # Resize image to target size
        img = io.resize(img, target_size, anti_aliasing=True)
        # Convert image to grayscale
        img = np.mean(img, axis=2)
        preprocessed_images.append(img)
    return preprocessed_images


def extract_features(images):
    features = []
    for img in images:
        # Compute image morphology features
        morph_features = []
        for i in range(5, 20, 5):
            selem = feature.disk(i)
            eroded = feature.erosion(img, selem)
            dilated = feature.dilation(img, selem)
            opening = feature.opening(img, selem)
            closing = feature.closing(img, selem)
            morph_features.extend([np.mean(eroded), np.mean(dilated), np.mean(opening), np.mean(closing)])
        # Compute image texture features
        texture_features = feature.greycomatrix(img, [1], [0, np.pi/4, np.pi/2, 3*np.pi/4], levels=256,
                                                 symmetric=True, normed=True)
        texture_features = np.ravel(texture_features)
        features.append(np.concatenate([morph_features, texture_features]))
    return features


def compare_distributions(feature1, feature2):
    _, p_value = kstest(feature1, feature2)
    return p_value


def plot_histogram(feature1, feature2, title):
    plt.hist(feature1, bins=50, alpha=0.5, label='Natural images')
    plt.hist(feature2, bins=50, alpha=0.5, label='Ultrasound images')
    plt.legend(loc='upper right')
    plt.title(title)
    plt.show()


# Load images
us_folder = 'D:/zzz/na5'
nat_folder = 'D:/zzz/us5'
us_images = load_images(us_folder)
nat_images = load_images(nat_folder)

# Preprocess images
target_size = (256, 256)
us_images = preprocess_images(us_images, target_size)
nat_images = preprocess_images(nat_images, target_size)

# Extract features
us_features = extract_features(us_images)
nat_features = extract_features(nat_images)

# Compare distributions of each feature
p_values = []
for i in range(len(us_features)):
    p_value = compare_distributions(us_features[i], nat_features[i])
    p_values.append(p_value)

# Plot histogram of p-values
plot_histogram(nat_features, us_features, 'Distribution of Morphology and Texture Features')
plt.hist(p_values, bins=50)
plt.title('Distribution of p-values')
plt.show()

# Perform statistical analysis
mean_p_value = np.mean(p_values)
std_p_value = np.std(p_values)
print('Mean p-value:', mean_p_value)
print('Standard deviation of p-values:', std_p_value)

# Visualize results
if mean_p_value < 0.05:
    print('There is a significant difference between the distributions of natural and ultrasound images.')
else:
    print('There is no significant difference between the distributions of natural and ultrasound images.')

此代码首先加载超声图像和自然图像文件夹中的所有图像，并将它们的大小调整为256x256像素。接下来，代码提取每个图像的形态和纹理特征，将这些特征存储到两个不同的特征向量中。对于每个特征，代码使用Kolmogorov-Smirnov检验比较自然图像和超声图像之间的分布。最后，代码绘制了每个特征的p值的直方图，以及平均p值和p值标准差的统计数据。如果平均p值小于0.05，则认为自然图像和超声图像之间存在显着差异。

希望这个代码能帮到你！