【618】不同图片格式的压缩情况

参考：图片格式分不清？JPG、TIF、PNG，不同格式区别到底在哪里

　　在制作深度学习语义分割的标签数据的时候，会涉及到存储图片的步骤，默认就是存成了 JPG 格式，今天跟同事聊天，突然说 JPG 会压缩数据，突然我也回忆道自己在生成的白色 JPG 图片上看到了莫名的淡灰色，于是自己打算自己试验下 JPG、PNG、TIF 与 BMP 数据是否能还原为原来的 numpy.array。结论就是 JPG 压缩很明显，其他数据格式都能还原，而且 PNG 占用的存储空间是最小的。

　　代码：

>>> import numpy as np
>>> from PIL import Image
>>> arr = np.arange(100).reshape(10, 10)
>>> arr = arr.astype("uint8")
>>> Image.fromarray(arr).save("1.jpg")
>>> Image.fromarray(arr).save("1.png")
>>> Image.fromarray(arr).save("1.tif")
>>> Image.fromarray(arr).save("1.bmp")
>>> img_jpg = Image.open("1.jpg")
>>> np.array(img_jpg)
array([[ 1,  1,  2,  3,  4,  5,  6,  7, 10, 10],
       [ 9, 10, 11, 12, 13, 14, 15, 15, 18, 18],
       [21, 21, 22, 23, 25, 26, 27, 27, 30, 30],
       [31, 31, 32, 33, 34, 36, 36, 37, 40, 40],
       [39, 40, 40, 42, 43, 44, 45, 45, 48, 48],
       [49, 49, 50, 51, 53, 54, 55, 55, 58, 58],
       [61, 61, 62, 63, 64, 65, 66, 67, 70, 70],
       [69, 70, 71, 72, 73, 74, 75, 75, 78, 78],
       [80, 81, 82, 83, 84, 85, 86, 86, 89, 89],
       [90, 90, 91, 92, 94, 95, 96, 96, 99, 99]], dtype=uint8)
>>> img_png = Image.open("1.png")
>>> np.array(img_png)
array([[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
       [20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
       [30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
       [40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
       [50, 51, 52, 53, 54, 55, 56, 57, 58, 59],
       [60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
       [70, 71, 72, 73, 74, 75, 76, 77, 78, 79],
       [80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
       [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]], dtype=uint8)
>>> img_tif = Image.open("1.tif")
>>> np.array(img_tif)
array([[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
       [20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
       [30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
       [40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
       [50, 51, 52, 53, 54, 55, 56, 57, 58, 59],
       [60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
       [70, 71, 72, 73, 74, 75, 76, 77, 78, 79],
       [80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
       [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]], dtype=uint8)
>>> img_bmp = Image.open("1.bmp")
>>> np.array(img_tif)
array([[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9],
       [10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
       [20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
       [30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
       [40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
       [50, 51, 52, 53, 54, 55, 56, 57, 58, 59],
       [60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
       [70, 71, 72, 73, 74, 75, 76, 77, 78, 79],
       [80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
       [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]], dtype=uint8)

　　图片大小：

　　结论就是优先存储为 PNG 格式。