用python实现基于概率分布的超声图像与自然图像性质差异分析，不用greycomatrix和greycoprops，使用opencv，两种图像的文件类型都是jpg,超声图像的文件路径是‘D:\zzz\us2’，自然图像的文件路径是‘D:\zzz\na2’，两种图像都有84张，图片形状都不同，图像名称是随机的，需要对比两种图像的特征必须有颜色，纹理，形状，差异性分析方法也需要多种，包括完整详细代码

以下是一种实现方式：

import cv2
import numpy as np
import os
import matplotlib.pyplot as plt

# 定义函数：计算颜色直方图
def calc_color_hist(image):
    hist = np.zeros((256, 3))
    for i in range(3):
        channel_hist = cv2.calcHist([image], [i], None, [256], [0,256])
        hist[:, i] = channel_hist[:, 0]
    return hist

# 定义函数：计算纹理特征（LBP）
def calc_texture_lbp(image):
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    lbp = np.zeros_like(gray)
    for i in range(1, gray.shape[0]-1):
        for j in range(1, gray.shape[1]-1):
            center = gray[i,j]
            code = 0
            code |= (gray[i-1,j-1] >= center) << 7
            code |= (gray[i-1,j] >= center) << 6
            code |= (gray[i-1,j+1] >= center) << 5
            code |= (gray[i,j+1] >= center) << 4
            code |= (gray[i+1,j+1] >= center) << 3
            code |= (gray[i+1,j] >= center) << 2
            code |= (gray[i+1,j-1] >= center) << 1
            code |= (gray[i,j-1] >= center) << 0
            lbp[i,j] = code
    hist = cv2.calcHist([lbp], [0], None, [256], [0,256])
    return hist

# 定义函数：计算形状特征（Hu矩）
def calc_shape_hu(image):
    gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    moments = cv2.moments(gray)
    hu_moments = cv2.HuMoments(moments)
    return hu_moments

# 定义函数：对比两张图片的颜色、纹理、形状特征差异
def compare_images(image1, image2):
    # 计算颜色直方图差异
    hist1 = calc_color_hist(image1)
    hist2 = calc_color_hist(image2)
    color_diff = cv2.compareHist(hist1, hist2, cv2.HISTCMP_CHISQR)

    # 计算纹理特征差异
    texture1 = calc_texture_lbp(image1)
    texture2 = calc_texture_lbp(image2)
    texture_diff = cv2.compareHist(texture1, texture2, cv2.HISTCMP_CHISQR)

    # 计算形状特征差异
    shape1 = calc_shape_hu(image1)
    shape2 = calc_shape_hu(image2)
    shape_diff = np.sum(np.abs(np.log(shape1) - np.log(shape2)))

    return color_diff, texture_diff, shape_diff

# 读取文件夹中的所有图片
us_dir = 'D:\\zzz\\us2'
na_dir = 'D:\\zzz\\na2'
us_images = []
na_images = []
for filename in os.listdir(us_dir):
    image = cv2.imread(os.path.join(us_dir, filename))
    us_images.append(image)
for filename in os.listdir(na_dir):
    image = cv2.imread(os.path.join(na_dir, filename))
    na_images.append(image)

# 计算每张超声图像与每张自然图像的特征差异
color_diffs = np.zeros((len(us_images), len(na_images)))
texture_diffs = np.zeros((len(us_images), len(na_images)))
shape_diffs = np.zeros((len(us_images), len(na_images)))
for i in range(len(us_images)):
    for j in range(len(na_images)):
        color_diff, texture_diff, shape_diff = compare_images(us_images[i], na_images[j])
        color_diffs[i, j] = color_diff
        texture_diffs[i, j] = texture_diff
        shape_diffs[i, j] = shape_diff

# 找到每张超声图像最相似的自然图像，输出结果
for i in range(len(us_images)):
    index = np.argmin(color_diffs[i,:] + texture_diffs[i,:] + shape_diffs[i,:])
    print("超声图像{}与自然图像{}最相似".format(i+1, index+1))

# 可视化结果：将每张超声图像与最相似的自然图像放在一起展示
for i in range(len(us_images)):
    index = np.argmin(color_diffs[i,:] + texture_diffs[i,:] + shape_diffs[i,:])
    plt.subplot(1, 2, 1)
    plt.imshow(cv2.cvtColor(us_images[i], cv2.COLOR_BGR2RGB))
    plt.title("超声图像{}".format(i+1))
    plt.xticks([])
    plt.yticks([])
    plt.subplot(1, 2, 2)
    plt.imshow(cv2.cvtColor(na_images[index], cv2.COLOR_BGR2RGB))
    plt.title("自然图像{}".format(index+1))
    plt.xticks([])
    plt.yticks([])
    plt.show()

该代码中实现了三种特征差异的计算方法：颜色直方图差异、LBP纹理特征差异、Hu矩形状特征差异。其中颜色直方图和LBP纹理特征使用OpenCV中的函数计算，而Hu矩需要自己实现。在对比两张图片时，将三种特征差异分别计算，并求和作为最终的差异值。通过比较每张超声图像与每张自然图像的差异值，找到最相似的自然图像，输出结果并可视化展示。