DnCNN Noise Reduction Image Noise Reduction Modeling Dissertation Writing Guide 10 Page Dissertation

profiledean990605
PMRID-Pytorch-main.zip
Home>Computer Science homework help>Python homework help>DnCNN Noise Reduction Image Noise Reduction Modeling Dissertation Writing Guide 10 Page Dissertation

__pycache__/Dncnn.cpython-38.pyc

cut.py

# import os # import SimpleITK as sitk # import numpy as np from PIL import Image # # # def slice_and_save_images(input_dir, output_dir, slice_size=64): # """ # 读取mhd+raw格式的医学图像,将其切片并保存为图片格式 # # 参数: # input_dir (str): 输入图像的目录路径 # output_dir (str): 输出切片图像的目录路径 # slice_size (int): 切片的大小 (默认为64x64) # """ # # 创建输出目录 # if not os.path.exists(output_dir): # os.makedirs(output_dir) # # # 遍历输入目录中的所有文件 # for filename in os.listdir(input_dir): # if filename.endswith('.mhd'): # # 读取mhd+raw格式的图像 # image_path = os.path.join(input_dir, filename) # image = sitk.ReadImage(image_path) # image_array = sitk.GetArrayFromImage(image) # # # 获取图像的维度 # depth, height, width = image_array.shape # # # 将图像切片并保存为图片格式 # for z in range(0, depth, slice_size): # for y in range(0, height, slice_size): # for x in range(0, width, slice_size): # slice_image = image_array[z:z + slice_size, y:y + slice_size, x:x + slice_size] # # # 将切片图像的数据类型转换为 8 位整型 # normalized_slice = (slice_image - slice_image.min()) / ( # slice_image.max() - slice_image.min()) * 255 # normalized_slice = normalized_slice.astype(np.uint8) # # slice_path = os.path.join(output_dir, f"slice_{z}_{y}_{x}.jpg") # sitk.WriteImage(sitk.GetImageFromArray(slice_image), slice_path) # print(f"Saved slice: {slice_path}") # # # # 示例用法 input_dir = r'C:\Users\Administrator\Desktop\data\TEST001' output_dir = r'./' # slice_and_save_images(input_dir, output_dir, slice_size=64) import os from tqdm import tqdm import SimpleITK as sitk import matplotlib.pyplot as plt import cv2 import shutil def slice(ori_path: str, pro_path: str): id = 0 for path in os.listdir(ori_path): if path.find('mhd') >= 0: id += 1 save_content = os.path.join(pro_path, str(id)) if os.path.exists(save_content): shutil.rmtree(save_content) os.makedirs(save_content) data_mhd = sitk.ReadImage(os.path.join(ori_path, path)) spacing = data_mhd.GetSpacing() scan = sitk.GetArrayFromImage(data_mhd) for i in tqdm(range(len(scan))): img = cv2.normalize(scan[i], None, 0, 255, cv2.NORM_MINMAX, cv2.CV_8U) img = Image.fromarray(img) save_path = os.path.join(save_content, f'{id}_{i}.png') img.save(save_path) slice(input_dir, output_dir)

Dncnn.py

import torch import torch.nn as nn import torch.nn.functional as F class DnCNN(nn.Module): def __init__(self, depth=17, n_channels=64, image_channels=1): super(DnCNN, self).__init__() kernel_size = 3 padding = 1 features = n_channels layers = [] layers.append(nn.Conv2d(in_channels=image_channels, out_channels=features, kernel_size=kernel_size, padding=padding, bias=True)) layers.append(nn.ReLU(inplace=True)) for _ in range(depth-2): layers.append(nn.Conv2d(in_channels=features, out_channels=features, kernel_size=kernel_size, padding=padding, bias=False)) layers.append(nn.BatchNorm2d(features)) layers.append(nn.ReLU(inplace=True)) layers.append(nn.Conv2d(in_channels=features, out_channels=image_channels, kernel_size=kernel_size, padding=padding, bias=False)) self.dncnn = nn.Sequential(*layers) def forward(self, x): out = self.dncnn(x) return out # Example usage if __name__ == "__main__": # Create a DnCNN model model = DnCNN(depth=17, n_channels=64, image_channels=1) # Load pre-trained weights (if available) # model.load_state_dict(torch.load('path/to/pretrained_model.pth')) # Move the model to the GPU (if available) device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu") model.to(device) # Prepare the input image input_image = torch.randn(1, 1, 256, 256).to(device) # Adjust the size as needed # Forward pass output = model(input_image) # Print the output shape print(output.shape)

generate_list.py

import os input_path = r"F:\earn\EDCNN-master\EDCNN-master\demo\1/" gt_path = r"F:\earn\EDCNN-master\EDCNN-master\demo\2/" list_path = "F:\earn\PMRID-Pytorch-main\PMRID-Pytorch-main/list" f = open(list_path, 'w') for root, tmp, files in os.walk(input_path, topdown=False): for name in files: name1='noisy_' + name if os.path.exists(gt_path+name1): input_file = input_path+name gt_file = gt_path+name1 print(input_file + " " + gt_file) f.writelines(input_file + " " + gt_file + " - - -" + "\n")

list

F:\earn\EDCNN-master\EDCNN-master\demo\1/1_0.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_0.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_1.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_1.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_10.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_10.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_100.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_100.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_101.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_101.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_102.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_102.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_103.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_103.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_104.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_104.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_105.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_105.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_106.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_106.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_107.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_107.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_108.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_108.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_109.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_109.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_11.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_11.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_110.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_110.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_111.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_111.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_112.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_112.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_113.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_113.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_114.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_114.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_115.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_115.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_116.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_116.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_117.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_117.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_118.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_118.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_119.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_119.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_12.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_12.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_120.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_120.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_121.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_121.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_122.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_122.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_123.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_123.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_124.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_124.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_125.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_125.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_126.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_126.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_127.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_127.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_13.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_13.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_14.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_14.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_15.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_15.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_16.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_16.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_17.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_17.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_18.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_18.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_19.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_19.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_2.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_2.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_20.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_20.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_21.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_21.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_22.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_22.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_23.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_23.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_24.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_24.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_25.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_25.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_26.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_26.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_27.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_27.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_28.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_28.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_29.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_29.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_3.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_3.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_30.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_30.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_31.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_31.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_32.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_32.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_33.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_33.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_34.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_34.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_35.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_35.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_36.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_36.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_37.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_37.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_38.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_38.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_39.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_39.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_4.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_4.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_40.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_40.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_41.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_41.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_42.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_42.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_43.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_43.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_44.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_44.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_45.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_45.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_46.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_46.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_47.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_47.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_48.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_48.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_49.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_49.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_5.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_5.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_50.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_50.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_51.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_51.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_52.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_52.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_53.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_53.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_54.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_54.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_55.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_55.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_56.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_56.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_57.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_57.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_58.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_58.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_59.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_59.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_6.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_6.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_60.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_60.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_61.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_61.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_62.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_62.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_63.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_63.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_64.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_64.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_65.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_65.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_66.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_66.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_67.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_67.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_68.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_68.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_69.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_69.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_7.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_7.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_70.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_70.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_71.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_71.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_72.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_72.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_73.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_73.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_74.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_74.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_75.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_75.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_76.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_76.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_77.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_77.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_78.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_78.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_79.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_79.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_8.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_8.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_80.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_80.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_81.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_81.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_82.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_82.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_83.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_83.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_84.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_84.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_85.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_85.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_86.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_86.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_87.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_87.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_88.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_88.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_89.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_89.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_9.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_9.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_90.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_90.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_91.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_91.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_92.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_92.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_93.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_93.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_94.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_94.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_95.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_95.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_96.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_96.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_97.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_97.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_98.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_98.png - - - F:\earn\EDCNN-master\EDCNN-master\demo\1/1_99.png F:\earn\EDCNN-master\EDCNN-master\demo\2/noisy_1_99.png - - -

logs/pmrid_l2/events.out.tfevents.1721904029.Administrator.32136.0

logs/pmrid_l2/events.out.tfevents.1721910184.Administrator.25772.0

logs/pmrid_l2/events.out.tfevents.1721910210.Administrator.35008.0

logs/pmrid_l2/events.out.tfevents.1721910291.Administrator.31576.0

logs/pmrid_l2/events.out.tfevents.1721910440.Administrator.41012.0

logs/pmrid_l2/events.out.tfevents.1721910472.Administrator.22532.0

logs/pmrid_l2/events.out.tfevents.1721910531.Administrator.32472.0

logs/pmrid_l2/events.out.tfevents.1721910697.Administrator.18800.0

logs/pmrid_l2/events.out.tfevents.1721910763.Administrator.40288.0

logs/pmrid_l2/events.out.tfevents.1721910888.Administrator.29528.0

logs/pmrid_l2/events.out.tfevents.1721910917.Administrator.15228.0

logs/pmrid_l2/events.out.tfevents.1721913471.Administrator.28048.0

logs/pmrid_l2/events.out.tfevents.1721916116.Administrator.41796.0

logs/pmrid_l2/events.out.tfevents.1721916604.Administrator.41240.0

logs/pmrid_l2/events.out.tfevents.1721917634.Administrator.12128.0

logs/pmrid_l2/events.out.tfevents.1721918161.Administrator.26620.0

logs/pmrid_l2/events.out.tfevents.1721919426.Administrator.28092.0

logs/pmrid_l2/events.out.tfevents.1721920750.Administrator.38528.0

logs/pmrid_l2/events.out.tfevents.1721920853.Administrator.7036.0

logs/pmrid_l2/events.out.tfevents.1721921473.Administrator.21240.0

logs/pmrid_l2/events.out.tfevents.1721956248.Administrator.14776.0

logs/pmrid_l2/events.out.tfevents.1721956265.Administrator.18468.0

logs/pmrid_l2/events.out.tfevents.1721958274.Administrator.5176.0

logs/pmrid_l2/events.out.tfevents.1721958890.Administrator.2668.0

logs/pmrid_l2/events.out.tfevents.1721960807.Administrator.4868.0

logs/pmrid_l2/events.out.tfevents.1721960831.Administrator.14088.0

logs/pmrid_l2/events.out.tfevents.1721962307.Administrator.152.0

logs/pmrid_l2/events.out.tfevents.1721965027.Administrator.16584.0

logs/pmrid_l2/events.out.tfevents.1721967388.Administrator.12544.0

main.py

from model.pmrid.pmrid_api import PMRID_API import torch if __name__=='__main__': pmrid_api = PMRID_API( 100, 4, 0.0001, 'cuda:0', './logs/pmrid_l2/', './params/pmrid_l2/', 'F:\earn\PMRID-Pytorch-main\PMRID-Pytorch-main\list', 'F:\earn\PMRID-Pytorch-main\PMRID-Pytorch-main\list', True, './model/pmrid/pmrid_pretrained.ckp' ) # pmrid_api.train_and_value() # train and value pmrid_api.test(r'F:\earn\PMRID-Pytorch-main\PMRID-Pytorch-main\params\pmrid_l2\91.ckp','./output/pmrid_l2_dataset/value/right/') # test

model/pmrid/__pycache__/pmrid.cpython-36.pyc

model/pmrid/__pycache__/pmrid.cpython-38.pyc

model/pmrid/__pycache__/pmrid_api.cpython-311.pyc

model/pmrid/__pycache__/pmrid_api.cpython-36.pyc

model/pmrid/__pycache__/pmrid_api.cpython-38.pyc

model/pmrid/__pycache__/utils.cpython-36.pyc

model/pmrid/__pycache__/utils.cpython-38.pyc

model/pmrid/pmrid.py

#!/usr/bin/env python3 import torch import torch.nn as nn from collections import OrderedDict import numpy as np def Conv2D( in_channels: int, out_channels: int, kernel_size: int, stride: int, padding: int, is_seperable: bool = False, has_relu: bool = False, ): modules = OrderedDict() if is_seperable: modules['depthwise'] = nn.Conv2d( in_channels, in_channels, kernel_size, stride, padding, groups=in_channels, bias=False, ) modules['pointwise'] = nn.Conv2d( in_channels, out_channels, kernel_size=1, stride=1, padding=0, bias=True, ) else: modules['conv'] = nn.Conv2d( in_channels, out_channels, kernel_size, stride, padding, bias=True, ) if has_relu: modules['relu'] = nn.ReLU() return nn.Sequential(modules) class EncoderBlock(nn.Module): def __init__(self, in_channels: int, mid_channels: int, out_channels: int, stride: int = 1): super().__init__() self.conv1 = Conv2D(in_channels, mid_channels, kernel_size=5, stride=stride, padding=2, is_seperable=True, has_relu=True) self.conv2 = Conv2D(mid_channels, out_channels, kernel_size=5, stride=1, padding=2, is_seperable=True, has_relu=False) self.proj = ( nn.Identity() if stride == 1 and in_channels == out_channels else Conv2D(in_channels, out_channels, kernel_size=3, stride=stride, padding=1, is_seperable=True, has_relu=False) ) self.relu = nn.ReLU() def forward(self, x): proj = self.proj(x) x = self.conv1(x) x = self.conv2(x) x = x + proj return self.relu(x) def EncoderStage(in_channels: int, out_channels: int, num_blocks: int): blocks = [ EncoderBlock( in_channels=in_channels, mid_channels=out_channels//4, out_channels=out_channels, stride=2, ) ] for _ in range(num_blocks-1): blocks.append( EncoderBlock( in_channels=out_channels, mid_channels=out_channels//4, out_channels=out_channels, stride=1, ) ) return nn.Sequential(*blocks) class DecoderBlock(nn.Module): def __init__(self, in_channels: int, out_channels: int, kernel_size: int = 3): super().__init__() padding = kernel_size // 2 self.conv0 = Conv2D( in_channels, out_channels, kernel_size=kernel_size, padding=padding, stride=1, is_seperable=True, has_relu=True, ) self.conv1 = Conv2D( out_channels, out_channels, kernel_size=kernel_size, padding=padding, stride=1, is_seperable=True, has_relu=False, ) def forward(self, x): inp = x x = self.conv0(x) x = self.conv1(x) x = x + inp return x class DecoderStage(nn.Module): def __init__(self, in_channels: int, skip_in_channels: int, out_channels: int): super().__init__() self.decode_conv = DecoderBlock(in_channels, in_channels, kernel_size=3) self.upsample = nn.ConvTranspose2d(in_channels, out_channels, kernel_size=2, stride=2, padding=0) self.proj_conv = Conv2D(skip_in_channels, out_channels, kernel_size=3, stride=1, padding=1, is_seperable=True, has_relu=True) # M.init.msra_normal_(self.upsample.weight, mode='fan_in', nonlinearity='linear') def forward(self, inputs): inp, skip = inputs x = self.decode_conv(inp) x = self.upsample(x) y = self.proj_conv(skip) return x + y class PMRID(nn.Module): def __init__(self): super().__init__() self.conv0 = Conv2D(in_channels=4, out_channels=16, kernel_size=3, padding=1, stride=1, is_seperable=False, has_relu=True) self.enc1 = EncoderStage(in_channels=16, out_channels=64, num_blocks=2) self.enc2 = EncoderStage(in_channels=64, out_channels=128, num_blocks=2) self.enc3 = EncoderStage(in_channels=128, out_channels=256, num_blocks=4) self.enc4 = EncoderStage(in_channels=256, out_channels=512, num_blocks=4) self.encdec = Conv2D(in_channels=512, out_channels=64, kernel_size=3, padding=1, stride=1, is_seperable=True, has_relu=True) self.dec1 = DecoderStage(in_channels=64, skip_in_channels=256, out_channels=64) self.dec2 = DecoderStage(in_channels=64, skip_in_channels=128, out_channels=32) self.dec3 = DecoderStage(in_channels=32, skip_in_channels=64, out_channels=32) self.dec4 = DecoderStage(in_channels=32, skip_in_channels=16, out_channels=16) self.out0 = DecoderBlock(in_channels=16, out_channels=16, kernel_size=3) self.out1 = Conv2D(in_channels=16, out_channels=4, kernel_size=3, stride=1, padding=1, is_seperable=False, has_relu=False) def forward(self, inp): conv0 = self.conv0(inp) conv1 = self.enc1(conv0) conv2 = self.enc2(conv1) conv3 = self.enc3(conv2) conv4 = self.enc4(conv3) conv5 = self.encdec(conv4) up3 = self.dec1((conv5, conv3)) up2 = self.dec2((up3, conv2)) up1 = self.dec3((up2, conv1)) x = self.dec4((up1, conv0)) x = self.out0(x) x = self.out1(x) pred = inp + x return pred if __name__ == "__main__": net = Network() img = torch.randn(1, 4, 64, 64, device=torch.device('cpu'), dtype=torch.float32) out = net(img)

model/pmrid/pmrid_api.py

import os import cv2 import skimage.metrics import numpy as np from typing import Tuple import torch import torch.optim as optim import torch.nn.functional as F from torch.utils.data import Dataset from torch.utils.data import DataLoader from torch.utils.tensorboard import SummaryWriter from model.pmrid.utils import RawUtils from model.pmrid.pmrid import PMRID class KSigma: def __init__(self, k_coeff: Tuple[float, float], b_coeff: Tuple[float, float, float], anchor: float, v: float = 959.0): self.K = np.poly1d(k_coeff) self.Sigma = np.poly1d(b_coeff) self.anchor = anchor self.v = v def __call__(self, img_01, iso: float, inverse=False): k, sigma = self.K(iso), self.Sigma(iso) k_a, sigma_a = self.K(self.anchor), self.Sigma(self.anchor) cvt_k = k_a / k cvt_b = (sigma / (k ** 2) - sigma_a / (k_a ** 2)) * k_a img = img_01 * self.v if not inverse: img = img * cvt_k + cvt_b else: img = (img - cvt_b) / cvt_k return img / self.v class DataProcess(): def __init__(self): self.k_sigma = KSigma( k_coeff=[0.0005995267, 0.00868861], b_coeff=[7.11772e-7, 6.514934e-4, 0.11492713], anchor=1600, ) def pre_process(self, bayer: np.ndarray, iso: float): # normalize bayer = bayer / 255.0 # bayer to rggb rggb = RawUtils.bayer2rggb(bayer) rggb = rggb.clip(0, 1) # padding H, W = rggb.shape[:2] ph, pw = (32-(H % 32))//2, (32-(W % 32))//2 self.ph, self.pw = ph, pw rggb = np.pad(rggb, [(ph, ph), (pw, pw), (0, 0)], 'constant') # transpose rggb = rggb.transpose(2, 0, 1) # ksigma rggb = self.k_sigma(rggb, iso) # inverse normalize rggb = rggb * 255.0 return rggb def post_process(self, rggb: np.ndarray, iso: float): # normalize rggb = rggb / 255.0 # ksigma rggb = self.k_sigma(rggb, iso, inverse = True) # transpose rggb = rggb.transpose(1, 2, 0) # inverse padding ph, pw = self.ph, self.pw rggb = rggb[ph:-ph, pw:-pw] # rggb to bayer bayer = RawUtils.rggb2bayer(rggb) bayer = bayer.clip(0, 1) # inverse normalize bayer = bayer * 255.0 return bayer class PMRIDDataset(Dataset): def __init__(self, filepath, data_process, train = False): self.input_path = [] self.gt_path = [] for line in open(filepath): self.input_path.append(line.split(" ")[0]) self.gt_path.append(line.split(" ")[1]) self.len = len(self.input_path) self.data_process = data_process def __getitem__(self, index): input_iso = 4300 gt_iso = 4300 input_bayer = cv2.imread(self.input_path[index], 0).astype(np.float32) input_rggb = self.data_process.pre_process(input_bayer, input_iso) input_data = torch.from_numpy(input_rggb) gt_bayer = cv2.imread(self.gt_path[index], 0).astype(np.float32) gt_rggb = self.data_process.pre_process(gt_bayer, gt_iso) gt_data = torch.from_numpy(gt_rggb) label = self.input_path[index].split('/')[-1] return input_data, gt_data, input_iso, gt_iso, label def __len__(self): return self.len class PMRID_API(): def __init__(self, epoch, batch_size, learning_rate, device, logs_path, params_path, train_list_path, value_list_path, is_load_pretrained, pretrained_path): # parameters self.epoch = epoch self.batch_size = batch_size self.learning_rate = learning_rate self.device = device self.logs_path = logs_path self.params_path = params_path self.train_list_path = train_list_path self.value_list_path = value_list_path self.is_load_pretrained = is_load_pretrained self.pretrained_path = pretrained_path # data process self.data_process = DataProcess() # data loader train_dataset = PMRIDDataset(self.train_list_path, self.data_process) value_dataset = PMRIDDataset(self.value_list_path, self.data_process) self.train_loader = DataLoader(train_dataset, shuffle=True, batch_size=batch_size) self.value_loader = DataLoader(value_dataset, shuffle=False, batch_size=1) # tensorboard if not os.path.exists(self.logs_path): os.makedirs(self.logs_path) self.writer = SummaryWriter(self.logs_path) # networks if not os.path.exists(self.params_path): os.makedirs(self.params_path) self.pmrid = PMRID() # self.criterion = torch.nn.L1Loss() self.criterion = torch.nn.MSELoss() self.optimizer = optim.Adam(self.pmrid.parameters(), lr=self.learning_rate) def load_weight(self, path): states = torch.load(path) self.pmrid.load_state_dict(states) self.pmrid.to(self.device) print('[load] load finished') def init_weight(self): pass def train(self, epoch): losses = [] for batch_idx, data in enumerate(self.train_loader, 0): inputs, gts, input_iso, gt_iso, label = data inputs, gts = inputs.to(self.device), gts.to(self.device) self.optimizer.zero_grad() outputs = self.pmrid(inputs) loss = self.criterion(outputs, gts) loss.backward() self.optimizer.step() losses.append(loss.item()) print('[train] epoch: %d, batch: %d, loss: %f' % (epoch + 1, batch_idx + 1, loss.item())) mean_loss = np.mean(losses) print('[value] epoch: %d, mean loss: %f' % (epoch + 1, mean_loss)) self.writer.add_scalar('loss', mean_loss, epoch + 1) def value(self, epoch): psnrs = [] ssims = [] with torch.no_grad(): for batch_idx, data in enumerate(self.value_loader, 0): inputs, gts, input_iso, gt_iso, label = data inputs = inputs.to(self.device) gts = inputs.to(self.device) # run pmrid outputs = self.pmrid(inputs) input_rggb = inputs.squeeze().cpu().numpy() input_bayer = self.data_process.post_process(input_rggb, input_iso[0]) / 255.0 gt_rggb = gts.squeeze().cpu().numpy() gt_bayer = self.data_process.post_process(gt_rggb, gt_iso[0]) / 255.0 output_rggb = outputs.squeeze().cpu().numpy() output_bayer = self.data_process.post_process(output_rggb, input_iso[0]) / 255.0 psnr = skimage.metrics.peak_signal_noise_ratio(gt_bayer, output_bayer) data_range = (0, 1) ssim = skimage.metrics.structural_similarity(gt_bayer, output_bayer, data_range=data_range) psnrs.append(float(psnr)) ssims.append(float(ssim)) print(f"input_bayer min: {input_bayer.min()}, max: {input_bayer.max()}") print(f"gt_bayer min: {gt_bayer.min()}, max: {gt_bayer.max()}") print(f"output_bayer min: {output_bayer.min()}, max: {output_bayer.max()}") print('[value] epoch: %d, batch: %d, pnsr: %f, ssim: %f' % (epoch + 1, batch_idx + 1, psnr, ssim)) mean_psnr = np.mean(psnrs) mean_ssim = np.mean(ssims) print(f"input_bayer min: {input_bayer.min()}, max: {input_bayer.max()}") print(f"gt_bayer min: {gt_bayer.min()}, max: {gt_bayer.max()}") print(f"output_bayer min: {output_bayer.min()}, max: {output_bayer.max()}") print('[value] epoch: %d, mean pnsr: %f, mean ssim: %f' % (epoch + 1, mean_psnr, mean_ssim)) self.writer.add_scalar('psnr', mean_psnr, epoch + 1) self.writer.add_scalar('ssim', mean_ssim, epoch + 1) def train_and_value(self): if self.is_load_pretrained: self.load_weight(self.pretrained_path) else: self.init_weight() for epoch in range(self.epoch): if self.is_load_pretrained: pass else: for param_group in self.optimizer.param_groups: param_group['lr'] = self.learning_rate * (0.5 ** (epoch // 20)) self.train(epoch) self.value(epoch) torch.save(self.pmrid.state_dict(), self.params_path+'/'+str(epoch+1)+'.ckp') def test(self, params_path, output_path): self.load_weight(params_path) with torch.no_grad(): for batch_idx, data in enumerate(self.value_loader, 0): inputs, gts, input_iso, gt_iso, label = data inputs = inputs.to(self.device) gts = inputs.to(self.device) # run pmrid outputs = self.pmrid(inputs) output_rggb = outputs.squeeze().cpu().numpy() output_bayer = self.data_process.post_process(output_rggb, input_iso[0]) if not os.path.exists(output_path): os.makedirs(output_path) print('[test] ' +output_path + label[0]) cv2.imwrite(output_path + label[0], output_bayer.astype(np.uint8)) if __name__=='__main__': pmrid_api = PMRID_API( 200, 10, 0.01, 'cuda:0', './logs/pmrid_l2/', './params/pmrid_l2/', './data/right_value_list.txt', './data/right_value_list.txt', True, './model/pmrid/pmrid_pretrained.ckp' ) pmrid_api.train_and_value()

model/pmrid/pmrid_pretrained.ckp

archive/data.pkl

archive/data/1965483200

archive/data/1965483296

archive/data/1990042896

archive/data/1990042992

archive/data/1999848880

archive/data/43483136

archive/data/81139184

archive/data/81213264

archive/data/81369568

archive/data/82205264

archive/data/82227056

archive/data/82426176

archive/data/82538720

archive/data/82559648

archive/data/82567504

archive/data/82694800

archive/data/82696528

archive/data/82703760

archive/data/82716240

archive/data/82769072

archive/data/82769328

archive/data/82773904

archive/data/82775680

archive/data/82800544

archive/data/82801056

archive/data/82801568

archive/data/82804000

archive/data/82875536

archive/data/82876432

archive/data/82890960

archive/data/82907984

archive/data/82908496

archive/data/82913424

archive/data/82930448

archive/data/82931344

archive/data/82945872

archive/data/82979280

archive/data/82979920

archive/data/82988048

archive/data/83054224

archive/data/83055632

archive/data/83061968

archive/data/83193680

archive/data/83195088

archive/data/83222416

archive/data/83288592

archive/data/83289232

archive/data/83297360

archive/data/83363536

archive/data/83364944

archive/data/83392272

archive/data/83458448

archive/data/83459088

archive/data/83467216

archive/data/83533392

archive/data/83534800

archive/data/83562128

archive/data/83628304

archive/data/83628944

archive/data/83637072

archive/data/83637584

archive/data/83638992

archive/data/83666320

archive/data/83666832

archive/data/83667728

archive/data/83682256

archive/data/83682768

archive/data/83685040

archive/data/83685200

archive/data/83686800

archive/data/83687312

archive/data/83689744

archive/data/83691344

archive/data/83691856

archive/data/83692752

archive/data/83694352

archive/data/83694864

archive/data/85278512

archive/data/85331504

archive/data/85594288

archive/data/85595184

archive/data/85609712

archive/data/85872496

archive/data/85874928

archive/data/85927856

archive/data/86190640

archive/data/86191536

archive/data/86206064

archive/data/86468848

archive/data/86471280

archive/data/86491440

archive/data/86623152

archive/data/86623792

archive/data/86628352

archive/data/86645920

archive/data/86666880

archive/data/86667520

archive/data/86746352

archive/data/86812528

archive/data/86813168

archive/data/86817200

archive/data/86834224

archive/data/86834864

archive/data/86838896

archive/data/86855920

archive/data/86856560

archive/data/86891056

archive/data/86891568

archive/data/86897488

archive/data/86914512

archive/data/86915024

archive/data/86917904

archive/data/86922640

archive/data/86923152

archive/data/86926032

archive/data/86930768

archive/data/86931280

archive/data/86949392

archive/data/86949904

archive/data/86953520

archive/data/86962352

archive/data/86962864

archive/data/86965744

archive/data/86970480

archive/data/86970992

archive/data/86973872

archive/data/86978608

archive/data/86979120

archive/data/86989040

archive/data/86989488

archive/data/86991376

archive/data/86993040

archive/data/86993488

archive/data/86995792

archive/data/86997456

archive/data/86997904

archive/data/87000208

archive/data/87001872

archive/data/87002320

archive/data/87006944

archive/data/87010448

archive/version

3

model/pmrid/utils.py

#!/usr/bin/env python3 import cv2 import numpy as np class RawUtils: @classmethod def bggr2rggb(cls, *bayers): res = [] for bayer in bayers: res.append(bayer[::-1, ::-1]) if len(res) == 1: return res[0] return res @classmethod def rggb2bggr(cls, *bayers): return cls.bggr2rggb(*bayers) @classmethod def bayer2rggb(cls, *bayers): res = [] for bayer in bayers: H, W = bayer.shape res.append( bayer.reshape(H//2, 2, W//2, 2) .transpose(0, 2, 1, 3) .reshape(H//2, W//2, 4) ) if len(res) == 1: return res[0] return res @classmethod def rggb2bayer(cls, *rggbs): res = [] for rggb in rggbs: H, W, _ = rggb.shape res.append( rggb.reshape(H, W, 2, 2) .transpose(0, 2, 1, 3) .reshape(H*2, W*2) ) if len(res) == 1: return res[0] return res @classmethod def bayer2rgb(cls, *bayer_01s, wb_gain, CCM, gamma=2.2): wb_gain = np.array(wb_gain)[[0, 1, 1, 2]] res = [] for bayer_01 in bayer_01s: bayer = cls.rggb2bayer( (cls.bayer2rggb(bayer_01) * wb_gain).clip(0, 1) ).astype(np.float32) bayer = np.round(np.ascontiguousarray(bayer) * 65535).clip(0, 65535).astype(np.uint16) rgb = cv2.cvtColor(bayer, cv2.COLOR_BAYER_BG2RGB_EA).astype(np.float32) / 65535 rgb = rgb.dot(np.array(CCM).T).clip(0, 1) rgb = rgb ** (1/gamma) res.append(rgb.astype(np.float32)) if len(res) == 1: return res[0] return res # vim: ts=4 sw=4 sts=4 expandtab

output/pmrid_l2_dataset/value/right/1_0.png

output/pmrid_l2_dataset/value/right/1_1.png

output/pmrid_l2_dataset/value/right/1_10.png

output/pmrid_l2_dataset/value/right/1_100.png

output/pmrid_l2_dataset/value/right/1_101.png

output/pmrid_l2_dataset/value/right/1_102.png

output/pmrid_l2_dataset/value/right/1_103.png

output/pmrid_l2_dataset/value/right/1_104.png

output/pmrid_l2_dataset/value/right/1_105.png

output/pmrid_l2_dataset/value/right/1_106.png

output/pmrid_l2_dataset/value/right/1_107.png

output/pmrid_l2_dataset/value/right/1_108.png

output/pmrid_l2_dataset/value/right/1_109.png

output/pmrid_l2_dataset/value/right/1_11.png

output/pmrid_l2_dataset/value/right/1_110.png

output/pmrid_l2_dataset/value/right/1_111.png

output/pmrid_l2_dataset/value/right/1_112.png

output/pmrid_l2_dataset/value/right/1_113.png

output/pmrid_l2_dataset/value/right/1_114.png

output/pmrid_l2_dataset/value/right/1_115.png

output/pmrid_l2_dataset/value/right/1_116.png

output/pmrid_l2_dataset/value/right/1_117.png

output/pmrid_l2_dataset/value/right/1_118.png

output/pmrid_l2_dataset/value/right/1_119.png

output/pmrid_l2_dataset/value/right/1_12.png

output/pmrid_l2_dataset/value/right/1_120.png

output/pmrid_l2_dataset/value/right/1_121.png

output/pmrid_l2_dataset/value/right/1_122.png

output/pmrid_l2_dataset/value/right/1_123.png

output/pmrid_l2_dataset/value/right/1_124.png

output/pmrid_l2_dataset/value/right/1_125.png

output/pmrid_l2_dataset/value/right/1_126.png

output/pmrid_l2_dataset/value/right/1_127.png

output/pmrid_l2_dataset/value/right/1_13.png

output/pmrid_l2_dataset/value/right/1_14.png

output/pmrid_l2_dataset/value/right/1_15.png

output/pmrid_l2_dataset/value/right/1_16.png

output/pmrid_l2_dataset/value/right/1_17.png

output/pmrid_l2_dataset/value/right/1_18.png

output/pmrid_l2_dataset/value/right/1_19.png

output/pmrid_l2_dataset/value/right/1_2.png

output/pmrid_l2_dataset/value/right/1_20.png

output/pmrid_l2_dataset/value/right/1_21.png

output/pmrid_l2_dataset/value/right/1_22.png

output/pmrid_l2_dataset/value/right/1_23.png

output/pmrid_l2_dataset/value/right/1_24.png

output/pmrid_l2_dataset/value/right/1_25.png

output/pmrid_l2_dataset/value/right/1_26.png

output/pmrid_l2_dataset/value/right/1_27.png

output/pmrid_l2_dataset/value/right/1_28.png

output/pmrid_l2_dataset/value/right/1_29.png

output/pmrid_l2_dataset/value/right/1_3.png

output/pmrid_l2_dataset/value/right/1_30.png

output/pmrid_l2_dataset/value/right/1_31.png

output/pmrid_l2_dataset/value/right/1_32.png

output/pmrid_l2_dataset/value/right/1_33.png

output/pmrid_l2_dataset/value/right/1_34.png

output/pmrid_l2_dataset/value/right/1_35.png

output/pmrid_l2_dataset/value/right/1_36.png

output/pmrid_l2_dataset/value/right/1_37.png

output/pmrid_l2_dataset/value/right/1_38.png

output/pmrid_l2_dataset/value/right/1_39.png

output/pmrid_l2_dataset/value/right/1_4.png

output/pmrid_l2_dataset/value/right/1_40.png

output/pmrid_l2_dataset/value/right/1_41.png

output/pmrid_l2_dataset/value/right/1_42.png

output/pmrid_l2_dataset/value/right/1_43.png

output/pmrid_l2_dataset/value/right/1_44.png

output/pmrid_l2_dataset/value/right/1_45.png

output/pmrid_l2_dataset/value/right/1_46.png

output/pmrid_l2_dataset/value/right/1_47.png

output/pmrid_l2_dataset/value/right/1_48.png

output/pmrid_l2_dataset/value/right/1_49.png

output/pmrid_l2_dataset/value/right/1_5.png

output/pmrid_l2_dataset/value/right/1_50.png

output/pmrid_l2_dataset/value/right/1_51.png

output/pmrid_l2_dataset/value/right/1_52.png

output/pmrid_l2_dataset/value/right/1_53.png

output/pmrid_l2_dataset/value/right/1_54.png

output/pmrid_l2_dataset/value/right/1_55.png

output/pmrid_l2_dataset/value/right/1_56.png

output/pmrid_l2_dataset/value/right/1_57.png

output/pmrid_l2_dataset/value/right/1_58.png

output/pmrid_l2_dataset/value/right/1_59.png

output/pmrid_l2_dataset/value/right/1_6.png

output/pmrid_l2_dataset/value/right/1_60.png

output/pmrid_l2_dataset/value/right/1_61.png

output/pmrid_l2_dataset/value/right/1_62.png

output/pmrid_l2_dataset/value/right/1_63.png

output/pmrid_l2_dataset/value/right/1_64.png

output/pmrid_l2_dataset/value/right/1_65.png

output/pmrid_l2_dataset/value/right/1_66.png

output/pmrid_l2_dataset/value/right/1_67.png

output/pmrid_l2_dataset/value/right/1_68.png

output/pmrid_l2_dataset/value/right/1_69.png

output/pmrid_l2_dataset/value/right/1_7.png

output/pmrid_l2_dataset/value/right/1_70.png

output/pmrid_l2_dataset/value/right/1_71.png

output/pmrid_l2_dataset/value/right/1_72.png

output/pmrid_l2_dataset/value/right/1_73.png

output/pmrid_l2_dataset/value/right/1_74.png

output/pmrid_l2_dataset/value/right/1_75.png

output/pmrid_l2_dataset/value/right/1_76.png

output/pmrid_l2_dataset/value/right/1_77.png

output/pmrid_l2_dataset/value/right/1_78.png

output/pmrid_l2_dataset/value/right/1_79.png

output/pmrid_l2_dataset/value/right/1_8.png

output/pmrid_l2_dataset/value/right/1_80.png

output/pmrid_l2_dataset/value/right/1_81.png

output/pmrid_l2_dataset/value/right/1_82.png

output/pmrid_l2_dataset/value/right/1_83.png

output/pmrid_l2_dataset/value/right/1_84.png

output/pmrid_l2_dataset/value/right/1_85.png

output/pmrid_l2_dataset/value/right/1_86.png

output/pmrid_l2_dataset/value/right/1_87.png

output/pmrid_l2_dataset/value/right/1_88.png

output/pmrid_l2_dataset/value/right/1_89.png

output/pmrid_l2_dataset/value/right/1_9.png

output/pmrid_l2_dataset/value/right/1_90.png

output/pmrid_l2_dataset/value/right/1_91.png

output/pmrid_l2_dataset/value/right/1_92.png

output/pmrid_l2_dataset/value/right/1_93.png

output/pmrid_l2_dataset/value/right/1_94.png

output/pmrid_l2_dataset/value/right/1_95.png

output/pmrid_l2_dataset/value/right/1_96.png

output/pmrid_l2_dataset/value/right/1_97.png

output/pmrid_l2_dataset/value/right/1_98.png

output/pmrid_l2_dataset/value/right/1_99.png

params/pmrid_l2/1.ckp

params/pmrid_l2/10.ckp

params/pmrid_l2/11.ckp

params/pmrid_l2/12.ckp

params/pmrid_l2/13.ckp

params/pmrid_l2/14.ckp

params/pmrid_l2/15.ckp

params/pmrid_l2/16.ckp

params/pmrid_l2/17.ckp

params/pmrid_l2/18.ckp

params/pmrid_l2/19.ckp

params/pmrid_l2/2.ckp

params/pmrid_l2/20.ckp

params/pmrid_l2/21.ckp

params/pmrid_l2/22.ckp

params/pmrid_l2/23.ckp

params/pmrid_l2/24.ckp

params/pmrid_l2/25.ckp

params/pmrid_l2/26.ckp

params/pmrid_l2/27.ckp

params/pmrid_l2/28.ckp

params/pmrid_l2/29.ckp

params/pmrid_l2/3.ckp

params/pmrid_l2/30.ckp

params/pmrid_l2/31.ckp

params/pmrid_l2/32.ckp

params/pmrid_l2/33.ckp

params/pmrid_l2/34.ckp

params/pmrid_l2/35.ckp

params/pmrid_l2/36.ckp

params/pmrid_l2/37.ckp

params/pmrid_l2/38.ckp

params/pmrid_l2/39.ckp

params/pmrid_l2/4.ckp

params/pmrid_l2/40.ckp

params/pmrid_l2/41.ckp

params/pmrid_l2/42.ckp

params/pmrid_l2/43.ckp

params/pmrid_l2/44.ckp

params/pmrid_l2/45.ckp

params/pmrid_l2/46.ckp

params/pmrid_l2/47.ckp

params/pmrid_l2/48.ckp

params/pmrid_l2/49.ckp

params/pmrid_l2/5.ckp

params/pmrid_l2/50.ckp

params/pmrid_l2/51.ckp

params/pmrid_l2/52.ckp

params/pmrid_l2/53.ckp

params/pmrid_l2/54.ckp

params/pmrid_l2/55.ckp

params/pmrid_l2/56.ckp

params/pmrid_l2/57.ckp

params/pmrid_l2/58.ckp

params/pmrid_l2/59.ckp

params/pmrid_l2/6.ckp

params/pmrid_l2/60.ckp

params/pmrid_l2/61.ckp

params/pmrid_l2/62.ckp

params/pmrid_l2/63.ckp

params/pmrid_l2/64.ckp

params/pmrid_l2/65.ckp

params/pmrid_l2/66.ckp

params/pmrid_l2/67.ckp

params/pmrid_l2/68.ckp

params/pmrid_l2/69.ckp

params/pmrid_l2/7.ckp

params/pmrid_l2/70.ckp

params/pmrid_l2/71.ckp

params/pmrid_l2/72.ckp

params/pmrid_l2/73.ckp

params/pmrid_l2/74.ckp

params/pmrid_l2/75.ckp

params/pmrid_l2/76.ckp

params/pmrid_l2/77.ckp

params/pmrid_l2/78.ckp

params/pmrid_l2/79.ckp

params/pmrid_l2/8.ckp

params/pmrid_l2/80.ckp

params/pmrid_l2/81.ckp

params/pmrid_l2/82.ckp

params/pmrid_l2/83.ckp

params/pmrid_l2/84.ckp

params/pmrid_l2/85.ckp

params/pmrid_l2/86.ckp

params/pmrid_l2/87.ckp

params/pmrid_l2/88.ckp

params/pmrid_l2/89.ckp

params/pmrid_l2/9.ckp

params/pmrid_l2/90.ckp

params/pmrid_l2/91.ckp

poit.py

import cv2 import os import numpy as np # Set the directory containing the images image_dir = "./1/" out = './2/' # Loop through all the files in the directory for filename in os.listdir(image_dir): # Check if the file is an image if filename.endswith(".png") or filename.endswith(".jpg"): # Read the image image_path = os.path.join(image_dir, filename) image = cv2.imread(image_path) # Set the amount of salt-and-pepper noise s_vs_p = 0.5 amount = 0.04 # Create a copy of the image noisy_img = np.copy(image) # Add salt noise num_salt = np.ceil(amount * image.size * s_vs_p) coords = [np.random.randint(0, i - 1, int(num_salt)) for i in image.shape] noisy_img[coords[0], coords[1], :] = [255, 255, 255] # Add pepper noise num_pepper = np.ceil(amount * image.size * (1. - s_vs_p)) coords = [np.random.randint(0, i - 1, int(num_pepper)) for i in image.shape] noisy_img[coords[0], coords[1], :] = [0, 0, 0] # Save the noisy image noisy_image_path = os.path.join(out, "noisy_" + filename) cv2.imwrite(noisy_image_path, noisy_img)

README.md

# PMRID-Pytorch this is a project for training and testing PMRID some code copy from [megvii-research/PMRID](https://github.com/megvii-research/PMRID) generate_list.py is for generate list for trainning main.py is for training and testing dataset

try.py

import cv2 import numpy as np # 读取原始图像和去噪后的图像 noisy_image = r'F:\earn\EDCNN-master\EDCNN-master\demo\2\noisy_1_0.png' noisy_image = cv2.imread(noisy_image) gray_image = cv2.cvtColor(noisy_image, cv2.COLOR_BGR2GRAY) denoised_image = cv2.medianBlur(gray_image, 5) # 5表示核的大小，可以根据需要调整 # 计算 PSNR psnr = cv2.PSNR(gray_image, denoised_image) print(f"PSNR: {psnr:.2f} dB") # 计算 SSIM def ssim(img1, img2): C1 = (0.01 * 255)**2 C2 = (0.03 * 255)**2 img1 = img1.astype(np.float64) img2 = img2.astype(np.float64) kernel = cv2.getGaussianKernel(11, 1.5) window = np.outer(kernel, kernel.transpose()) mu1 = cv2.filter2D(img1, -1, window)[5:-5, 5:-5] mu2 = cv2.filter2D(img2, -1, window)[5:-5, 5:-5] mu1_sq = mu1**2 mu2_sq = mu2**2 mu1_mu2 = mu1 * mu2 sigma1_sq = cv2.filter2D(img1**2, -1, window)[5:-5, 5:-5] - mu1_sq sigma2_sq = cv2.filter2D(img2**2, -1, window)[5:-5, 5:-5] - mu2_sq sigma12 = cv2.filter2D(img1 * img2, -1, window)[5:-5, 5:-5] - mu1_mu2 ssim_map = ((2 * mu1_mu2 + C1) * (2 * sigma12 + C2)) / ((mu1_sq + mu2_sq + C1) * (sigma1_sq + sigma2_sq + C2)) return ssim_map.mean() ssim_score = ssim(gray_image, denoised_image) print(f"SSIM: {ssim_score:.4f}") # 显示图像 cv2.imshow('Original Image', noisy_image) cv2.imshow('Denoised Image', denoised_image) cv2.waitKey(0) cv2.destroyAllWindows() # import os # import cv2 # import numpy as np # import torch # import torch.nn as nn # import torch.optim as optim # from torch.utils.data import Dataset, DataLoader # from Dncnn import DnCNN # # # class DenoisingDataset(Dataset): # def __init__(self, data_dir, patch_size=40, stride=10): # self.data_dir = data_dir # self.patch_size = patch_size # self.stride = stride # self.file_list = os.listdir(data_dir) # # def __len__(self): # return len(self.file_list) # # def __getitem__(self, idx): # file_name = self.file_list[idx] # img_path = os.path.join(self.data_dir, file_name) # img = cv2.imread(img_path, cv2.IMREAD_GRAYSCALE) # # # 从原始图像中提取小块patch # patches = [] # for i in range(0, img.shape[0] - self.patch_size + 1, self.stride): # for j in range(0, img.shape[1] - self.patch_size + 1, self.stride): # patch = img[i:i + self.patch_size, j:j + self.patch_size] # patches.append(patch) # # # 将patch转换为Tensor # patches = np.stack(patches, axis=0) # patches = torch.from_numpy(patches).float().unsqueeze(1) # # # 添加高斯噪声 # noise = torch.randn_like(patches) * 25 # noisy_patches = patches + noise # # return noisy_patches, patches # # def train_dncnn(dataset, model, device, num_epochs=100, batch_size=128, lr=0.001): # dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True, num_workers=4) # criterion = nn.MSELoss() # optimizer = optim.Adam(model.parameters(), lr=lr) # # for epoch in range(num_epochs): # for i, (noisy_patches, clean_patches) in enumerate(dataloader): # noisy_patches = noisy_patches.to(device) # clean_patches = clean_patches.to(device) # # optimizer.zero_grad() # output = model(noisy_patches) # loss = criterion(output, clean_patches) # loss.backward() # optimizer.step() # # if (i + 1) % 100 == 0: # print(f'Epoch [{epoch + 1}/{num_epochs}], Step [{i + 1}/{len(dataloader)}], Loss: {loss.item():.4f}') # # return model # # # 创建 DnCNN 模型 # model = DnCNN(depth=17, n_channels=64, image_channels=1) # device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu") # model.to(device) # # # 创建数据集和训练模型 # dataset = DenoisingDataset(r'F:\earn\EDCNN-master\EDCNN-master\demo\2') # trained_model = train_dncnn(dataset, model, device) # # # 保存训练好的模型 # torch.save(trained_model.state_dict(), 'F:\earn\PMRID-Pytorch-main\PMRID-Pytorch-main/model.pth')