python提取图像的内角直方图特征

提取图像的内角直方图特征可以通过以下步骤实现：

1. 读取图像并进行灰度化处理。

import cv2

img = cv2.imread('image.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

2. 对图像进行角点检测。

corners = cv2.goodFeaturesToTrack(gray, 100, 0.01, 10)
corners = np.int0(corners)

3. 计算角点的内角度数。

angles = []
for i in corners:
    x, y = i.ravel()
    for j in corners:
        if i[0][0] != j[0][0] or i[0][1] != j[0][1]:
            x1, y1 = j.ravel()
            angle = np.arctan2(y1 - y, x1 - x) * 180 / np.pi
            if angle < 0:
                angle += 360
            angles.append(angle)

4. 统计角度分布并画直方图。

hist, bins = np.histogram(angles, bins=10, range=(0, 360))
plt.hist(angles, bins=10, range=(0, 360))
plt.xlabel('Angle (degrees)')
plt.ylabel('Frequency')
plt.title('Inner Corner Angle Histogram')
plt.show()

完整代码如下：

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

img = cv2.imread('image.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

corners = cv2.goodFeaturesToTrack(gray, 100, 0.01, 10)
corners = np.int0(corners)

angles = []
for i in corners:
    x, y = i.ravel()
    for j in corners:
        if i[0][0] != j[0][0] or i[0][1] != j[0][1]:
            x1, y1 = j.ravel()
            angle = np.arctan2(y1 - y, x1 - x) * 180 / np.pi
            if angle < 0:
                angle += 360
            angles.append(angle)

hist, bins = np.histogram(angles, bins=10, range=(0, 360))
plt.hist(angles, bins=10, range=(0, 360))
plt.xlabel('Angle (degrees)')
plt.ylabel('Frequency')
plt.title('Inner Corner Angle Histogram')
plt.show()