728x90
이미지 파일을 npy 로 변환하고 불러오기¶
In [ ]:
## 이미지 파일을 npy 파일로 변환
In [ ]:
import os
from PIL import Image
import numpy as np
import matplotlib.pyplot as plt
def load_images_to_numpy(folder_path):
# 폴더 내의 모든 파일 이름을 가져옵니다
file_names = [f for f in os.listdir(folder_path) if os.path.isfile(os.path.join(folder_path, f))]
# 이미지를 저장할 빈 리스트를 생성합니다
images = []
# 각 파일에 대해 반복
for file_name in file_names:
file_path = os.path.join(folder_path, file_name)
# 이미지를 열고 numpy 배열로 변환
with Image.open(file_path) as img:
images.append(np.array(img))
# 모든 이미지를 하나의 numpy 배열로 합칩니다
all_images_array = np.array(images)
return all_images_array
# 폴더 경로 설정
folder_path = './pokemon'
# 이미지 로드 및 numpy 배열로 변환
sprites = load_images_to_numpy(folder_path)
# 결과 확인
print(sprites)
# 배열을 .npy 파일로 저장
np.save('sprites.npy', sprites)
[[[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]] [[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]] [[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]] ... [[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]] [[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]] [[[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] ... [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]] [[0 0 0 0] [0 0 0 0] [0 0 0 0] ... [0 0 0 0] [0 0 0 0] [0 0 0 0]]]]
In [ ]:
sprites = np.load('sprites.npy')
sprites
Out[ ]:
array([[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]],
[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]],
[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]],
...,
[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]],
[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]],
[[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
...,
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]],
[[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0],
...,
[0, 0, 0, 0],
[0, 0, 0, 0],
[0, 0, 0, 0]]]], dtype=uint8)
In [ ]:
# 준비된 (64,64)크기의 나비 이미지를 (28,28)로 리사이즈하고
# 픽셀값을 0~1로 노멀라이즈
x0 = (sprites / 255)[12]
# 노멀라이즈 확인
print(x0.min(), x0.max())
0.0 1.0
In [ ]:
plt.imshow(x0)
plt.show()
In [ ]:
728x90
'공부정리 > Computer Vision' 카테고리의 다른 글
| [핵심 머신러닝] Class Activation Map (CAM) (1) (0) | 2024.01.15 |
|---|---|
| [Diffusion] 간단한 diffusion 모델로 포켓몬 학습하기 (0) | 2024.01.12 |
| 생성모델 평가지표 - IS, FID (0) | 2024.01.12 |
| Denoisiong Diffusion Probabilistic Models (2) DDPM Loss Function (0) | 2024.01.12 |
| Denoisiong Diffusion Probabilistic Models (1) Diffusion model's Forward & Reverse Process (0) | 2024.01.12 |