I now have a batch of images that I want to convert to numpy data to save, but the image size is not fixed (I don"t want resize)
for example, a. Shapee = (3rec 2jue 9), b. Shapee = (3pr 3pr 8), how to merge a meme b into c? That is to say, to realize c [0] = a ~ ~ c [1] = b?
so I would like to ask, is there any way to save these inconsistent pictures in the form of numpy.array?
use the pillow module to convert all kinds of pictures into pixel sets, then convert them to numpy arrays, and finally save them to a file.
Please refer to the following code
-sharp -*- coding: utf-8 -*-
from PIL import Image
import numpy as np
def images_to_array(image_files, array_file):
""" numpy .npy
"""
data = []
for filename in image_files:
data.append(np.array(Image.open(filename)))
np.save(array_file, data)
def load_images(array_file):
""" .npy
"""
return np.load(array_file)
images_to_array(['1.png', '2.png'], '1.npy')
load_images('1.npy')references
how to query data in B dataframe according to the ID, of A dataframe in pandas, An and B tables have the same ID column. Similar to SQL s column-level subquery: select * from t_class where c_id in (select c_class_id from t_student) , An and B have an o...