Question

我有很多块文件，例如：

chunks_64x64_NORMvsDISTRESS_train_chunk_000_x.npy
chunks_64x64_NORMvsDISTRESS_train_chunk_000_y.npy

由于这些文件是图像分区和标签。如何将这些文件合并为一个文件？以及如何将它们转换为png或jpeg文件？

from gaps_dataset import gaps
import zipfile
import os
import numpy as np
import cv2

# change these parameters:
destination_directory = 'D:/local/datasets/gaps/v1'
login = 'login'

# download images and patch references
#gaps.download_images(login=login, datadir=destination_directory)

# unzip images
#zip_filename = os.path.join(destination_directory, 'images/images.zip')
#zip_ref = zipfile.ZipFile(zip_filename, 'r')
#zip_ref.extractall(os.path.join(destination_directory, 'images'))
#zip_ref.close()

# load patch references for training subset
ref_filename = os.path.join(destination_directory, 'train/patch_references_train.npy')
patch_ref = np.load(ref_filename).astype(int)

# get patches from images
n_patches = patch_ref.shape[0] # number of patches
image_template = os.path.join(destination_directory, 'images/images/train_{:04d}.jpg')
for patch_index, patch_info in enumerate(patch_ref):
    # get information for this patch
    image_index, row, col, mirror_state, binary_label, class_label = patch_info
    # load image containing this patch
    image_filename = image_template.format(image_index)
    image = cv2.imread(image_filename)
    # extract patch from image
    patch = image[row:row+64, col:col+64]
    # mirroring
    if mirror_state == 0: # flip rows
        patch = patch[::-1, :]
    elif mirror_state == 1: # flip cols
        patch = patch[:, ::-1]
    # binary_label:
    #    0 = intact road,
    #    1 = distress
    # class_label:
    #    0 = intact road,
    #    1 = applied patch,
    #    2 = pothole,
    #    3 = inlaid patch,
    #    4 = open joint,
    #    5 = crack

# replace train with valid or test respectively to extract patches of the other subsets

我希望输出是图像文件，但出现错误消息：

not enough values to unpack (expected 6, got 1)

Answer 1

您正在以下行中打开patch_info的包装：

image_index, row, col, mirror_state, binary_label, class_label = patch_info

但是补丁信息不知道有6个值，而是只有一个值。请调试您的代码，并检查patch_info的值。

如何加入2个文件.npy并转换为图像文件？

1 个答案: