改进的极值中值滤波程序

极值中值滤波是一种数字信号处理技术，用于去除信号中的噪声。它通过计算信号的中值和极值来确定噪声的存在并进行滤波。以下是一个改进的极值中值滤波程序：

import numpy as np
from scipy import signal

def improved_peaks_median_filter(signal, window_size=3, threshold=0.5):
    """
    Improved peaks median filter algorithm
    :param signal: input signal
    :param window_size: size of the window
    :param threshold: threshold for detecting peaks
    :return: filtered signal
    """
    # Apply a median filter to the signal
    median_filtered = signal.medfilt(signal, window_size)
    
    # Compute the difference between the original signal and the median filtered signal
    diff_signal = np.abs(signal - median_filtered)
    
    # Compute the mean and standard deviation of the difference signal
    mean_diff = np.mean(diff_signal)
    std_diff = np.std(diff_signal)
    
    # Detect peaks in the difference signal
    peaks, _ = signal.find_peaks(diff_signal, height=(mean_diff + threshold * std_diff))
    
    # Replace the values at the detected peaks with the median filtered values
    filtered_signal = np.copy(signal)
    for peak in peaks:
        filtered_signal[peak] = median_filtered[peak]
        
    return filtered_signal

该算法首先对信号应用中值滤波器，然后计算信号的绝对值与中值滤波器输出之间的差异。它通过计算差异信号的平均值和标准差来检测噪声峰值，并将大于阈值的峰值替换为中值滤波器的输出。最后，它返回滤波信号。

你可以调整窗口大小和峰值阈值来适应不同的信号和噪声类型。