【热图生成】关键点检测/信号定位，heatmap生成

代码参考chatgpt生成

import numpy as np
import matplotlib.pyplot as plt
from scipy.ndimage import gaussian_filter
import random
# 假设我们有一些关键点位置
keypoints = [(50, 50), (70, 80), (90, 30)]

# 创建一个空白的热图
heatmap = np.zeros((100, 100))

# 将关键点位置添加到热图中
for point in keypoints:
    y, x = point
    heatmap[y, x] = random.random()
    # heatmap[y, x] = 1  # 假设置信度为1

# 使用高斯滤波平滑热图
heatmap_smooth = gaussian_filter(heatmap, sigma=1)

# 可视化热图
plt.imshow(heatmap_smooth, cmap='hot', interpolation='nearest')
plt.show()
print(heatmap_smooth.shape)

效果展示：heatmap[y, x] = random.random()决定每个点概率

heatmap_smooth = gaussian_filter(heatmap, sigma=1)决定整体热力大小
修改 sigma=3，效果如下: