dataset.py

目前觉得图像预处理是一件挺难的事情，有蛮多复杂的事情。

今天还是分析DnCNN两个版本的代码

https://github.com/SaoYan/DnCNN-PyTorch/blob/master/dataset.py

def Im2Patch(img, win, stride=1):
    k = 0
    endc = img.shape[0]
    endw = img.shape[1]
    endh = img.shape[2]
    patch = img[:, 0:endw-win+0+1:stride, 0:endh-win+0+1:stride]
    TotalPatNum = patch.shape[1] * patch.shape[2]
    Y = np.zeros([endc, win*win,TotalPatNum], np.float32)
    for i in range(win):
        for j in range(win):
            patch = img[:,i:endw-win+i+1:stride,j:endh-win+j+1:stride]
            Y[:,k,:] = np.array(patch[:]).reshape(endc, TotalPatNum)
            k = k + 1
    return Y.reshape([endc, win, win, TotalPatNum])

顾名思义，将一整张图片划分成一个个小的patch。img.shape通道数，宽、高。老实说，这段代码，我不是很能理解。。。希望有大神带带我

def gen_patches(file_name):
    # get multiscale patches from a single image
    img = cv2.imread(file_name, 0)  # gray scale
    h, w = img.shape
    patches = []
    for s in scales:
        h_scaled, w_scaled = int(h*s), int(w*s)
        img_scaled = cv2.resize(img, (h_scaled, w_scaled), interpolation=cv2.INTER_CUBIC)
        # extract patches
        for i in range(0, h_scaled-patch_size+1, stride):
            for j in range(0, w_scaled-patch_size+1, stride):
                x = img_scaled[i:i+patch_size, j:j+patch_size]
                for k in range(0, aug_times):
                    x_aug = data_aug(x, mode=np.random.randint(0, 8))
                    patches.append(x_aug)
    return patches

这个提取patch我能理解 i是每个patch开始的地方，每隔步长有一个 i+patch_size提取的部分。

下一部分是准备数据，代码如下，版本一：

def prepare_data(data_path, patch_size, stride, aug_times=1):
    # train
    print('process training data')
    scales = [1, 0.9, 0.8, 0.7]
    files = glob.glob(os.path.join(data_path, 'train', '*.png'))
    files.sort()
    h5f = h5py.File('train.h5', 'w')
    train_num = 0
    for i in range(len(files)):
        img = cv2.imread(files[i])
        h, w, c = img.shape
        for k in range(len(scales)):
            Img = cv2.resize(img, (int(h*scales[k]), int(w*scales[k])), interpolation=cv2.INTER_CUBIC)
            Img = np.expand_dims(Img[:,:,0].copy(), 0)
            Img = np.float32(normalize(Img))
            patches = Im2Patch(Img, win=patch_size, stride=stride)
            print("file: %s scale %.1f # samples: %d" % (files[i], scales[k], patches.shape[3]*aug_times))
            for n in range(patches.shape[3]):
                data = patches[:,:,:,n].copy()
                h5f.create_dataset(str(train_num), data=data)
                train_num += 1
                for m in range(aug_times-1):
                    data_aug = data_augmentation(data, np.random.randint(1,8))
                    h5f.create_dataset(str(train_num)+"_aug_%d" % (m+1), data=data_aug)
                    train_num += 1
    h5f.close()

后面还有一部分是验证集的我就不粘过来了。

scales = [1, 0.9, 0.8, 0.7]  缩放的尺度，为了有更多的数据。

files = glob.glob(os.path.join(data_path, 'train', '*.png'))  这句话还挺重要的：看教程https://blog.csdn.net/zmdzbzbhss123/article/details/52279008

通过cv2.resize实现缩放功能，expand_dims增加一个维度，具体见教程 https://www.jianshu.com/p/da10840660cb 但是不明白为什么要这样做，但是在所有关于数据集的预处理上都会有这种操作。

normalize解释https://blog.csdn.net/weicao1990/article/details/54584788?utm_source=blogxgwz6

关于h5的知识一直是空白的，但是一般图像处理完了都会保存在这里面，现在恶补一下相关知识：http://docs.h5py.org/en/stable/quick.html#quick

https://blog.csdn.net/y_f_raquelle/article/details/84134851

class Dataset(udata.Dataset):
    def __init__(self, train=True):
        super(Dataset, self).__init__()
        self.train = train
        if self.train:
            h5f = h5py.File('train.h5', 'r')
        else:
            h5f = h5py.File('val.h5', 'r')
        self.keys = list(h5f.keys())
        random.shuffle(self.keys)
        h5f.close()
    def __len__(self):
        return len(self.keys)
    def __getitem__(self, index):
        if self.train:
            h5f = h5py.File('train.h5', 'r')
        else:
            h5f = h5py.File('val.h5', 'r')
        key = self.keys[index]
        data = np.array(h5f[key])
        h5f.close()
        return torch.Tensor(data)

最后还定义了一个类，鉴于我是半路出家，并不能理解class是怎么回事儿，所以我可能要先学习一下类，这样才能活学活用吧