图像的离散小波分解（不同种类噪声，Python）

大家好，欢迎来到IT知识分享网。

import matplotlib.image as mpimg import matplotlib.pyplot as plt import numpy as np import sys import os import cv2 from skimage.util import random_noise sys.path.append('../') from Wavelet_2D import Wavelet_2D SAVE_OUNTPUTS = False if SAVE_OUNTPUTS: OUTPUT_DIR = './Outputs' OUTPUT_FORMAT = '.pdf' if not os.path.isdir(OUTPUT_DIR): os.mkdir(OUTPUT_DIR) plt.rc('text', usetex=True) plt.rc('font', family='serif') image = mpimg.imread('IMG__014553_621.png') image = image[:, :, :3] if image.dtype != np.uint8: tmp1 = np.zeros_like(image, dtype=np.uint8) for j in range(3): tmp1[:, :, j] = 255 * (image[:, :, j] - np.min(image[:, :, j])) / (np.max(image[:, :, j]) - np.min(image[:, :, j])) image = tmp1 image = cv2.resize(image, (256, 256)) image_sp = 255*random_noise(image, mode='s&p', amount=0.1) image_gs = 255*random_noise(image, mode='gaussian', mean=0, var=0.05) plt.figure(figsize=(10,5)) plt.subplot(1, 3, 1) plt.imshow(image) plt.title('Original Image', fontsize=15) plt.xticks([]); plt.yticks([]) plt.subplot(1, 3, 2) plt.imshow(image_sp.astype(np.uint8())) plt.title('Salt\&Pepper Noise', fontsize=15) plt.xticks([]); plt.yticks([]) plt.subplot(1, 3, 3) plt.imshow(image_gs.astype(np.uint8())) plt.title('Gaussian Noise', fontsize=15) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Image'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() 

SS_haar, SD_haar, DS_haar, DD_haar = Wavelet_2D(image,'Haar', 5) SS_haar_sp, SD_haar_sp, DS_haar_sp, DD_haar_sp = Wavelet_2D(image_sp,'Haar', 5) SS_haar_gs, SD_haar_gs, DS_haar_gs, DD_haar_gs = Wavelet_2D(image_gs,'Haar', 5)

 Processing Haar Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Haar Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Haar Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed!

SS_db4, SD_db4, DS_db4, DD_db4 = Wavelet_2D(image,'db4', 5) SS_db4_sp, SD_db4_sp, DS_db4_sp, DD_db4_sp = Wavelet_2D(image_sp,'db4', 5) SS_db4_gs, SD_db4_gs, DS_db4_gs, DD_db4_gs = Wavelet_2D(image_gs,'db4', 5)

 Processing Daubechies4 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Daubechies4 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Daubechies4 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed!

SS_db6, SD_db6, DS_db6, DD_db6 = Wavelet_2D(image,'db6', 5) SS_db6_sp, SD_db6_sp, DS_db6_sp, DD_db6_sp = Wavelet_2D(image_sp,'db6', 5) SS_db6_gs, SD_db6_gs, DS_db6_gs, DD_db6_gs = Wavelet_2D(image_gs,'db6', 5)

 Processing Daubechies6 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Daubechies6 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Daubechies6 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed!

SS_mh, SD_mh, DS_mh, DD_mh = Wavelet_2D(image,'mexicanhat', 5) SS_mh_sp, SD_mh_sp, DS_mh_sp, DD_mh_sp = Wavelet_2D(image_sp,'mexicanhat', 5) SS_mh_gs, SD_mh_gs, DS_mh_gs, DD_mh_gs = Wavelet_2D(image_gs,'mexicanhat', 5)

 Processing Mexican Hat Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Mexican Hat Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Mexican Hat Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed!

SS_sym2, SD_sym2, DS_sym2, DD_sym2 = Wavelet_2D(image,'sym2', 5) SS_sym2_sp, SD_sym2_sp, DS_sym2_sp, DD_sym2_sp = Wavelet_2D(image_sp,'sym2', 5) SS_sym2_gs, SD_sym2_gs, DS_sym2_gs, DD_sym2_gs = Wavelet_2D(image_gs,'sym2', 5)

 Processing Symlet2 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Symlet2 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed! Processing Symlet2 Wavelet Decomposition ... Decomposition Level 1 Completed! Decomposition Level 2 Completed! Decomposition Level 3 Completed! Decomposition Level 4 Completed! Decomposition Level 5 Completed!

import matplotlib.pyplot as plt plt.rc('text', usetex=True) plt.rc('font', family='serif') plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Haar Wavelet - Image Without Noise', fontsize=20) for i in range(len(SS_haar)): tmp1 = SS_haar[i].astype(np.uint8()) tmp2 = SD_haar[i].astype(np.uint8()) tmp3 = DS_haar[i].astype(np.uint8()) tmp4 = DD_haar[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_HaarWavelet_WithoutNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_sp.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Haar Wavelet - Image With S\&P Noise', fontsize=20) for i in range(len(SS_haar_sp)): tmp1 = SS_haar_sp[i].astype(np.uint8()) tmp2 = SD_haar_sp[i].astype(np.uint8()) tmp3 = DS_haar_sp[i].astype(np.uint8()) tmp4 = DD_haar_sp[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_HaarWavelet_SPNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_gs.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Haar Wavelet - Image With Gaussian Noise', fontsize=20) for i in range(len(SS_haar_gs)): tmp1 = SS_haar_gs[i].astype(np.uint8()) tmp2 = SD_haar_gs[i].astype(np.uint8()) tmp3 = DS_haar_gs[i].astype(np.uint8()) tmp4 = DD_haar_gs[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_HaarWavelet_GaussianNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show()

import matplotlib.pyplot as plt plt.rc('text', usetex=True) plt.rc('font', family='serif') plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies4 Wavelet - Image Without Noise', fontsize=20) for i in range(len(SS_db4)): tmp1 = SS_db4[i].astype(np.uint8()) tmp2 = SD_db4[i].astype(np.uint8()) tmp3 = DS_db4[i].astype(np.uint8()) tmp4 = DD_db4[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies4Wavelet_WithoutNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_sp.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies4 Wavelet - Image With S\&P Noise', fontsize=20) for i in range(len(SS_db4_sp)): tmp1 = SS_db4_sp[i].astype(np.uint8()) tmp2 = SD_db4_sp[i].astype(np.uint8()) tmp3 = DS_db4_sp[i].astype(np.uint8()) tmp4 = DD_db4_sp[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies4Wavelet_SPNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_gs.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies4 Wavelet - Image With Gaussian Noise', fontsize=20) for i in range(len(SS_db4_gs)): tmp1 = SS_db4_gs[i].astype(np.uint8()) tmp2 = SD_db4_gs[i].astype(np.uint8()) tmp3 = DS_db4_gs[i].astype(np.uint8()) tmp4 = DD_db4_gs[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies4Wavelet_GaussianNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show()

import matplotlib.pyplot as plt plt.rc('text', usetex=True) plt.rc('font', family='serif') plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies6 Wavelet - Image Without Noise', fontsize=20) for i in range(len(SS_db6)): tmp1 = SS_db6[i].astype(np.uint8()) tmp2 = SD_db6[i].astype(np.uint8()) tmp3 = DS_db6[i].astype(np.uint8()) tmp4 = DD_db6[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies6Wavelet_WithoutNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_sp.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies6 Wavelet - Image With S\&P Noise', fontsize=20) for i in range(len(SS_db6_sp)): tmp1 = SS_db6_sp[i].astype(np.uint8()) tmp2 = SD_db6_sp[i].astype(np.uint8()) tmp3 = DS_db6_sp[i].astype(np.uint8()) tmp4 = DD_db6_sp[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies6Wavelet_SPNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_gs.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Daubechies6 Wavelet - Image With Gaussian Noise', fontsize=20) for i in range(len(SS_db6_gs)): tmp1 = SS_db6_gs[i].astype(np.uint8()) tmp2 = SD_db6_gs[i].astype(np.uint8()) tmp3 = DS_db6_gs[i].astype(np.uint8()) tmp4 = DD_db6_gs[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Daubechies6Wavelet_GaussianNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show()

import matplotlib.pyplot as plt plt.rc('text', usetex=True) plt.rc('font', family='serif') plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Mexican Hat Wavelet - Image Without Noise', fontsize=20) for i in range(len(SS_mh)): tmp1 = SS_mh[i].astype(np.uint8()) tmp2 = SD_mh[i].astype(np.uint8()) tmp3 = DS_mh[i].astype(np.uint8()) tmp4 = DD_mh[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_MexicanHatWavelet_WithoutNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_sp.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Mexican Hat Wavelet - Image With S\&P Noise', fontsize=20) for i in range(len(SS_mh_sp)): tmp1 = SS_mh_sp[i].astype(np.uint8()) tmp2 = SD_mh_sp[i].astype(np.uint8()) tmp3 = DS_mh_sp[i].astype(np.uint8()) tmp4 = DD_mh_sp[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_MexicanHatWavelet_SPNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_gs.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Mexican Hat Wavelet - Image With Gaussian Noise', fontsize=20) for i in range(len(SS_mh_gs)): tmp1 = SS_mh_gs[i].astype(np.uint8()) tmp2 = SD_mh_gs[i].astype(np.uint8()) tmp3 = DS_mh_gs[i].astype(np.uint8()) tmp4 = DD_mh_gs[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_MexicanHatWavelet_GaussianNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show()

import matplotlib.pyplot as plt plt.rc('text', usetex=True) plt.rc('font', family='serif') plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Symlet2 Wavelet - Image Without Noise', fontsize=20) for i in range(len(SS_sym2)): tmp1 = SS_sym2[i].astype(np.uint8()) tmp2 = SD_sym2[i].astype(np.uint8()) tmp3 = DS_sym2[i].astype(np.uint8()) tmp4 = DD_sym2[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Symlet2Wavelet_WithoutNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_sp.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Symlet2 Wavelet - Image With S\&P Noise', fontsize=20) for i in range(len(SS_sym2_sp)): tmp1 = SS_sym2_sp[i].astype(np.uint8()) tmp2 = SD_sym2_sp[i].astype(np.uint8()) tmp3 = DS_sym2_sp[i].astype(np.uint8()) tmp4 = DD_sym2_sp[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Symlet2Wavelet_SPNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show() plt.figure(figsize=(12, 8)) plt.subplot(2, 3, 1) plt.imshow(image_gs.astype(np.uint8())) plt.title('Input') plt.xticks([]); plt.yticks([]) plt.suptitle('Symlet2 - Image With Gaussian Noise', fontsize=20) for i in range(len(SS_sym2_gs)): tmp1 = SS_sym2_gs[i].astype(np.uint8()) tmp2 = SD_sym2_gs[i].astype(np.uint8()) tmp3 = DS_sym2_gs[i].astype(np.uint8()) tmp4 = DD_sym2_gs[i].astype(np.uint8()) tmp11 = np.vstack((tmp1, tmp2)) tmp22 = np.vstack((tmp3, tmp4)) tmp = np.hstack((tmp11, tmp22)) plt.subplot(2, 3, i+2) plt.title(f'Decomposition Level: {i+1}') plt.imshow(tmp.astype(np.uint8())) plt.xticks([]); plt.yticks([]) plt.tight_layout() if SAVE_OUNTPUTS: plt.savefig(OUTPUT_DIR+'/2D_Symlet2Wavelet_GaussianNoise'+OUTPUT_FORMAT, bbox_inches='tight') plt.show()

知乎学术咨询：

https://www.zhihu.com/consult/people/?isMe=1

担任《Mechanical System and Signal Processing》《中国电机工程学报》等期刊审稿专家，擅长领域：信号滤波/降噪，机器学习/深度学习，时间序列预分析/预测，设备故障诊断/缺陷检测/异常检测。

分割线分割线

基于改进高斯-拉普拉斯滤波器的微震信号平滑降噪方法（MATLAB）

完整数据可通过知乎学术咨询获得

https://www.zhihu.com/consult/people/

非平稳信号的凸一维全变差降噪方法（MATLAB）

算法可迁移至金融时间序列，地震/微震信号，机械振动信号，声发射信号，电压/电流信号，语音信号，声信号，生理信号（ECG,EEG,EMG）等信号。

完整数据可通过知乎学术咨询获得

https://www.zhihu.com/consult/people/

基于广义交叉验证阈值的微震信号降噪方法（MATLAB）

算法可迁移至金融时间序列，地震/微震信号，机械振动信号，声发射信号，电压/电流信号，语音信号，声信号，生理信号（ECG,EEG,EMG）等信号。

完整数据可通过知乎学术咨询获得

https://www.zhihu.com/consult/people/

集合高阶统计量和小波块阈值的非平稳信号降噪方法-以地震信号为例（MATLAB）

完整数据可通过知乎学术咨询获得

https://www.zhihu.com/consult/people/ 

一种新的类谱峭度算法的旋转机械故障诊断模型（Python）

完整数据可通过知乎学术咨询获得

https://www.zhihu.com/consult/people/

采用8种方法对一维信号进行降噪（Python）

• NS: noisy signal
• S: original siganl
• mean filter: ws = window size
• median filter:
• average filter: ns = number of noisy signal(different)
• bandpass filter: l = low cut-off frequency, h = high …
• threshold filter: r = ratio(max abs(fft) / min …)
• wavelet filter: a = threshold
• std filter:
• NN: neural network

完整代码：

https://mbd.pub/o/bread/mbd-ZpyTmJZ