python+连续小波变换（CWT）

import librosa
import matplotlib.pyplot as plt
import numpy as np
import pywt
import librosa.display
wav, sr_ret = librosa.load('E:\PycharmProjects\pythonProject\AudioClassification-Pytorch-master\dataset/audio/fold1/1_001.wav',sr=48000,duration=0.1)
# 连续小波变换
wavename = 'morl'
totalscal = 4
fc = pywt.central_frequency(wavename)  # 计算小波函数的中心频率
cparam = 2 * fc * totalscal  # 常数c
scales = cparam / np.arange(totalscal, 1, -1)  # 为使转换后的频率序列是一等差序列，尺度序列必须取为这一形式（也即小波尺度）
[cwtmatr, frequencies] = pywt.cwt(wav, scales, wavename, 1.0 / sr_ret)
# scatter = Scattering2D(cwtmatr,frequencies)
# features =(scatter,wav)
# print(features.shape)
features = np.array(cwtmatr)
# frames = np.array(frames)
# print(features.shape)
# features = plt.contourf(t, frequencies, abs(cwtmatr))
# features = librosa.stft(y=wav, n_fft=1024, hop_length=None, win_length=None, window='hann')
# features = librosa.amplitude_to_db(np.abs(librosa.stft(y=wav,n_fft=1024, hop_length=None, win_length=None, window='hann')), ref=np.max)
# features = librosa.melspectrogram(y=wav,sr=sr, n_fft=400, n_mels=80, hop_length=160, win_length=400)
# features = librosa.power_to_db(features, ref=1.0, amin=1e-10, top_db=None)
# 归一化
mean = np.mean(features, 0, keepdims=True)
std = np.std(features, 0, keepdims=True)
features = (features - mean) / (std + 1e-5)
features = features.astype('float32')
print(features)