比较各种不同取样方式的图像缩放效果。
[NEAREST, BILINEAR, BICUBIC, LANCZOS, BOX, HAMMING]
NEAREST取样方式是效果最差的,PIL.Image.resize默认的resample方式就是使用NEAREST
import os from PIL import Image from PIL.Image import NEAREST, BILINEAR, BICUBIC, LANCZOS, BOX, HAMMING resmaple_list = [NEAREST, BILINEAR, BICUBIC, LANCZOS, BOX, HAMMING] path = r'C:\Users\UserName\Desktop\Wallpapers' filename = 'Biosphere Montreal.jpg' fullname = os.path.join(path, filename) image = Image.open(fullname) cut_size = [int(x*0.1) for x in image.size] new_size = (cut_size[0]*3, cut_size[1]*2) new_image = Image.new('RGB', new_size) for i in range(0, 3): im = image.resize(cut_size, resample=resmaple_list[i]) new_image.paste(im, (i*cut_size[0], 0)) for i in range(3, 6): im = image.resize(cut_size, resample=resmaple_list[i]) new_image.paste(im, ((i-3)*cut_size[0], cut_size[1])) new_image.show() new_image.save(os.path.join(path, 'outputImage.jpg'))
结果图:
原图左上角细节如下: