OpenCV 将图像读取为 3 通道图像，而 PIL 将图像读取为 1 通道图像
[英]OpenCV reads an image as a 3-channel image while PIL reads the same image as a 1-channel image
问题描述
I'm trying to figure out the NYU-v2 for semantic segmentation dataset in this Repository .
我试图在这个Repository 中找出语义分割数据集的 NYU-v2。 I'm having a hard time trying to understand how the image labels have been stored.我很难理解图像标签是如何存储的。
For example, given the following image :例如，给定以下图像：
The corresponding label image is:对应的标签图片为：
Now if I open the label image in OpenCV it reads this image as an RGB image:(wrongly, though. Even tried -1 as the second argument in imread )现在，如果我在 OpenCV 中打开标签图像，它会将这个图像读取为 RGB 图像:(不过是错误的。甚至尝试将 -1 作为imread的第二个参数）
import cv2
import numpy as np
img = img = cv2.imread(labelImage)
img.shape
>>> (468, 625, 3)
np.unque(img)
>>> array([ 0, 64, 128, 192, 224], dtype=uint8)
Whereas, if I open this same file in PIL:然而，如果我在 PIL 中打开同一个文件：
from PIL import Image
import numpy as np
img = np.array(Image.open(labelImage))
img.shape
>>> (468, 625)
np.unique(i)
>>> array([ 0, 2, 4, 13, 25, 39, 255], dtype=uint8)
Now the result that PIL shows is correct, and np.unique confirms this as each image in the NYU dataset has labels ranging from 0-39(255 is ignored), where each object is given a label in the 2-d space.现在 PIL 显示的结果是正确的，并且np.unique证实了这一点，因为纽约大学数据集中的每个图像的标签范围为 0-39（255 被忽略），其中每个对象在np.unique空间中都有一个标签。
Now the real question.现在是真正的问题。 How is the author storing the images such that they show up as color images in my OS's image viewer and OpenCV, but show up as single-channel images in PIL?作者如何存储图像，使其在我的操作系统的图像查看器和 OpenCV 中显示为彩色图像，但在 PIL 中显示为单通道图像？
1 个解决方案
解决方案1
10 已采纳 2020-01-21 11:09:16
It is stored as a palette (indexed) image.它存储为调色板（索引）图像。 See here .见这里。 OpenCV doesn't really like palette images and just converts to RGB. OpenCV并不真正喜欢调色板图像，只是转换为 RGB。
If I dump it with ImageMagick in Terminal, you can see the palette (under Colormap ) and the 7 colours under Histogram :如果我在终端中使用ImageMagick转储它，您可以看到调色板（在Colormap下）和Histogram下的 7 种颜色：
identify -verbose YourImage.png
Sample Output样本输出
Image: YourImage.png
Format: PNG (Portable Network Graphics)
Mime type: image/png
Class: PseudoClass
Geometry: 625x468+0+0
Units: Undefined
Colorspace: sRGB
Type: Palette
Base type: Undefined
Endianess: Undefined
Depth: 8-bit
Channel depth:
Red: 8-bit
Green: 8-bit
Blue: 8-bit
Channel statistics:
Pixels: 292500
Red:
min: 0 (0)
max: 224 (0.878431)
mean: 64.0019 (0.250988)
standard deviation: 87.3558 (0.342572)
kurtosis: -1.23408
skewness: 0.749894
entropy: 0.746999
Green:
min: 0 (0)
max: 224 (0.878431)
mean: 37.0024 (0.145107)
standard deviation: 68.3573 (0.268068)
kurtosis: 1.38689
skewness: 1.64292
entropy: 0.594363
Blue:
min: 0 (0)
max: 192 (0.752941)
mean: 61.0934 (0.239582)
standard deviation: 80.7328 (0.316599)
kurtosis: -1.34383
skewness: 0.676587
entropy: 0.842137
Image statistics:
Overall:
min: 0 (0)
max: 224 (0.878431)
mean: 54.0326 (0.211892)
standard deviation: 78.8153 (0.30908)
kurtosis: -0.752325
skewness: 0.974955
entropy: 0.727833
Colors: 7
Histogram:
163911: ( 0, 0, 0) #000000 black
16626: ( 0, 0,128) #000080 navy
5261: ( 0,128, 0) #008000 green
37681: (128,128,192) #8080C0 srgb(128,128,192)
36210: (192, 0,128) #C00080 srgb(192,0,128)
12644: (192, 64, 0) #C04000 srgb(192,64,0)
20167: (224,224,192) #E0E0C0 srgb(224,224,192)
Colormap entries: 256
Colormap:
0: ( 0, 0, 0,255) #000000FF black
1: (128, 0, 0,255) #800000FF maroon
2: ( 0,128, 0,255) #008000FF green
3: (128,128, 0,255) #808000FF olive
4: ( 0, 0,128,255) #000080FF navy
5: (128, 0,128,255) #800080FF purple
6: ( 0,128,128,255) #008080FF teal
7: (128,128,128,255) #808080FF fractal
8: ( 64, 0, 0,255) #400000FF srgba(64,0,0,1)
9: (192, 0, 0,255) #C00000FF srgba(192,0,0,1)
10: ( 64,128, 0,255) #408000FF srgba(64,128,0,1)
11: (192,128, 0,255) #C08000FF srgba(192,128,0,1)
12: ( 64, 0,128,255) #400080FF srgba(64,0,128,1)
13: (192, 0,128,255) #C00080FF srgba(192,0,128,1)
14: ( 64,128,128,255) #408080FF srgba(64,128,128,1)
15: (192,128,128,255) #C08080FF srgba(192,128,128,1)
16: ( 0, 64, 0,255) #004000FF srgba(0,64,0,1)
17: (128, 64, 0,255) #804000FF srgba(128,64,0,1)
18: ( 0,192, 0,255) #00C000FF srgba(0,192,0,1)
19: (128,192, 0,255) #80C000FF srgba(128,192,0,1)
20: ( 0, 64,128,255) #004080FF srgba(0,64,128,1)
21: (128, 64,128,255) #804080FF srgba(128,64,128,1)
22: ( 0,192,128,255) #00C080FF srgba(0,192,128,1)
23: (128,192,128,255) #80C080FF srgba(128,192,128,1)
24: ( 64, 64, 0,255) #404000FF srgba(64,64,0,1)
25: (192, 64, 0,255) #C04000FF srgba(192,64,0,1)
26: ( 64,192, 0,255) #40C000FF srgba(64,192,0,1)
27: (192,192, 0,255) #C0C000FF srgba(192,192,0,1)
28: ( 64, 64,128,255) #404080FF srgba(64,64,128,1)
29: (192, 64,128,255) #C04080FF srgba(192,64,128,1)
30: ( 64,192,128,255) #40C080FF srgba(64,192,128,1)
31: (192,192,128,255) #C0C080FF srgba(192,192,128,1)
32: ( 0, 0, 64,255) #000040FF srgba(0,0,64,1)
33: (128, 0, 64,255) #800040FF srgba(128,0,64,1)
34: ( 0,128, 64,255) #008040FF srgba(0,128,64,1)
35: (128,128, 64,255) #808040FF srgba(128,128,64,1)
36: ( 0, 0,192,255) #0000C0FF srgba(0,0,192,1)
37: (128, 0,192,255) #8000C0FF srgba(128,0,192,1)
38: ( 0,128,192,255) #0080C0FF srgba(0,128,192,1)
39: (128,128,192,255) #8080C0FF srgba(128,128,192,1)
40: ( 64, 0, 64,255) #400040FF srgba(64,0,64,1)
41: (192, 0, 64,255) #C00040FF srgba(192,0,64,1)
42: ( 64,128, 64,255) #408040FF srgba(64,128,64,1)
43: (192,128, 64,255) #C08040FF srgba(192,128,64,1)
44: ( 64, 0,192,255) #4000C0FF srgba(64,0,192,1)
45: (192, 0,192,255) #C000C0FF srgba(192,0,192,1)
46: ( 64,128,192,255) #4080C0FF srgba(64,128,192,1)
47: (192,128,192,255) #C080C0FF srgba(192,128,192,1)
48: ( 0, 64, 64,255) #004040FF srgba(0,64,64,1)
49: (128, 64, 64,255) #804040FF srgba(128,64,64,1)
50: ( 0,192, 64,255) #00C040FF srgba(0,192,64,1)
51: (128,192, 64,255) #80C040FF srgba(128,192,64,1)
52: ( 0, 64,192,255) #0040C0FF srgba(0,64,192,1)
53: (128, 64,192,255) #8040C0FF srgba(128,64,192,1)
54: ( 0,192,192,255) #00C0C0FF srgba(0,192,192,1)
55: (128,192,192,255) #80C0C0FF srgba(128,192,192,1)
56: ( 64, 64, 64,255) #404040FF grey25
57: (192, 64, 64,255) #C04040FF srgba(192,64,64,1)
58: ( 64,192, 64,255) #40C040FF srgba(64,192,64,1)
59: (192,192, 64,255) #C0C040FF srgba(192,192,64,1)
60: ( 64, 64,192,255) #4040C0FF srgba(64,64,192,1)
61: (192, 64,192,255) #C040C0FF srgba(192,64,192,1)
62: ( 64,192,192,255) #40C0C0FF srgba(64,192,192,1)
63: (192,192,192,255) #C0C0C0FF silver
64: ( 32, 0, 0,255) #200000FF srgba(32,0,0,1)
65: (160, 0, 0,255) #A00000FF srgba(160,0,0,1)
66: ( 32,128, 0,255) #208000FF srgba(32,128,0,1)
67: (160,128, 0,255) #A08000FF srgba(160,128,0,1)
68: ( 32, 0,128,255) #200080FF srgba(32,0,128,1)
69: (160, 0,128,255) #A00080FF srgba(160,0,128,1)
70: ( 32,128,128,255) #208080FF srgba(32,128,128,1)
71: (160,128,128,255) #A08080FF srgba(160,128,128,1)
72: ( 96, 0, 0,255) #600000FF srgba(96,0,0,1)
73: (224, 0, 0,255) #E00000FF srgba(224,0,0,1)
74: ( 96,128, 0,255) #608000FF srgba(96,128,0,1)
75: (224,128, 0,255) #E08000FF srgba(224,128,0,1)
76: ( 96, 0,128,255) #600080FF srgba(96,0,128,1)
77: (224, 0,128,255) #E00080FF srgba(224,0,128,1)
78: ( 96,128,128,255) #608080FF srgba(96,128,128,1)
79: (224,128,128,255) #E08080FF srgba(224,128,128,1)
80: ( 32, 64, 0,255) #204000FF srgba(32,64,0,1)
81: (160, 64, 0,255) #A04000FF srgba(160,64,0,1)
82: ( 32,192, 0,255) #20C000FF srgba(32,192,0,1)
83: (160,192, 0,255) #A0C000FF srgba(160,192,0,1)
84: ( 32, 64,128,255) #204080FF srgba(32,64,128,1)
85: (160, 64,128,255) #A04080FF srgba(160,64,128,1)
86: ( 32,192,128,255) #20C080FF srgba(32,192,128,1)
87: (160,192,128,255) #A0C080FF srgba(160,192,128,1)
88: ( 96, 64, 0,255) #604000FF srgba(96,64,0,1)
89: (224, 64, 0,255) #E04000FF srgba(224,64,0,1)
90: ( 96,192, 0,255) #60C000FF srgba(96,192,0,1)
91: (224,192, 0,255) #E0C000FF srgba(224,192,0,1)
92: ( 96, 64,128,255) #604080FF srgba(96,64,128,1)
93: (224, 64,128,255) #E04080FF srgba(224,64,128,1)
94: ( 96,192,128,255) #60C080FF srgba(96,192,128,1)
95: (224,192,128,255) #E0C080FF srgba(224,192,128,1)
96: ( 32, 0, 64,255) #200040FF srgba(32,0,64,1)
97: (160, 0, 64,255) #A00040FF srgba(160,0,64,1)
98: ( 32,128, 64,255) #208040FF srgba(32,128,64,1)
99: (160,128, 64,255) #A08040FF srgba(160,128,64,1)
100: ( 32, 0,192,255) #2000C0FF srgba(32,0,192,1)
101: (160, 0,192,255) #A000C0FF srgba(160,0,192,1)
102: ( 32,128,192,255) #2080C0FF srgba(32,128,192,1)
103: (160,128,192,255) #A080C0FF srgba(160,128,192,1)
104: ( 96, 0, 64,255) #600040FF srgba(96,0,64,1)
105: (224, 0, 64,255) #E00040FF srgba(224,0,64,1)
106: ( 96,128, 64,255) #608040FF srgba(96,128,64,1)
107: (224,128, 64,255) #E08040FF srgba(224,128,64,1)
108: ( 96, 0,192,255) #6000C0FF srgba(96,0,192,1)
109: (224, 0,192,255) #E000C0FF srgba(224,0,192,1)
110: ( 96,128,192,255) #6080C0FF srgba(96,128,192,1)
111: (224,128,192,255) #E080C0FF srgba(224,128,192,1)
112: ( 32, 64, 64,255) #204040FF srgba(32,64,64,1)
113: (160, 64, 64,255) #A04040FF srgba(160,64,64,1)
114: ( 32,192, 64,255) #20C040FF srgba(32,192,64,1)
115: (160,192, 64,255) #A0C040FF srgba(160,192,64,1)
116: ( 32, 64,192,255) #2040C0FF srgba(32,64,192,1)
117: (160, 64,192,255) #A040C0FF srgba(160,64,192,1)
118: ( 32,192,192,255) #20C0C0FF srgba(32,192,192,1)
119: (160,192,192,255) #A0C0C0FF srgba(160,192,192,1)
120: ( 96, 64, 64,255) #604040FF srgba(96,64,64,1)
121: (224, 64, 64,255) #E04040FF srgba(224,64,64,1)
122: ( 96,192, 64,255) #60C040FF srgba(96,192,64,1)
123: (224,192, 64,255) #E0C040FF srgba(224,192,64,1)
124: ( 96, 64,192,255) #6040C0FF srgba(96,64,192,1)
125: (224, 64,192,255) #E040C0FF srgba(224,64,192,1)
126: ( 96,192,192,255) #60C0C0FF srgba(96,192,192,1)
127: (224,192,192,255) #E0C0C0FF srgba(224,192,192,1)
128: ( 0, 32, 0,255) #002000FF srgba(0,32,0,1)
129: (128, 32, 0,255) #802000FF srgba(128,32,0,1)
130: ( 0,160, 0,255) #00A000FF srgba(0,160,0,1)
131: (128,160, 0,255) #80A000FF srgba(128,160,0,1)
132: ( 0, 32,128,255) #002080FF srgba(0,32,128,1)
133: (128, 32,128,255) #802080FF srgba(128,32,128,1)
134: ( 0,160,128,255) #00A080FF srgba(0,160,128,1)
135: (128,160,128,255) #80A080FF srgba(128,160,128,1)
136: ( 64, 32, 0,255) #402000FF srgba(64,32,0,1)
137: (192, 32, 0,255) #C02000FF srgba(192,32,0,1)
138: ( 64,160, 0,255) #40A000FF srgba(64,160,0,1)
139: (192,160, 0,255) #C0A000FF srgba(192,160,0,1)
140: ( 64, 32,128,255) #402080FF srgba(64,32,128,1)
141: (192, 32,128,255) #C02080FF srgba(192,32,128,1)
142: ( 64,160,128,255) #40A080FF srgba(64,160,128,1)
143: (192,160,128,255) #C0A080FF srgba(192,160,128,1)
144: ( 0, 96, 0,255) #006000FF srgba(0,96,0,1)
145: (128, 96, 0,255) #806000FF srgba(128,96,0,1)
146: ( 0,224, 0,255) #00E000FF srgba(0,224,0,1)
147: (128,224, 0,255) #80E000FF srgba(128,224,0,1)
148: ( 0, 96,128,255) #006080FF srgba(0,96,128,1)
149: (128, 96,128,255) #806080FF srgba(128,96,128,1)
150: ( 0,224,128,255) #00E080FF srgba(0,224,128,1)
151: (128,224,128,255) #80E080FF srgba(128,224,128,1)
152: ( 64, 96, 0,255) #406000FF srgba(64,96,0,1)
153: (192, 96, 0,255) #C06000FF srgba(192,96,0,1)
154: ( 64,224, 0,255) #40E000FF srgba(64,224,0,1)
155: (192,224, 0,255) #C0E000FF srgba(192,224,0,1)
156: ( 64, 96,128,255) #406080FF srgba(64,96,128,1)
157: (192, 96,128,255) #C06080FF srgba(192,96,128,1)
158: ( 64,224,128,255) #40E080FF srgba(64,224,128,1)
159: (192,224,128,255) #C0E080FF srgba(192,224,128,1)
160: ( 0, 32, 64,255) #002040FF srgba(0,32,64,1)
161: (128, 32, 64,255) #802040FF srgba(128,32,64,1)
162: ( 0,160, 64,255) #00A040FF srgba(0,160,64,1)
163: (128,160, 64,255) #80A040FF srgba(128,160,64,1)
164: ( 0, 32,192,255) #0020C0FF srgba(0,32,192,1)
165: (128, 32,192,255) #8020C0FF srgba(128,32,192,1)
166: ( 0,160,192,255) #00A0C0FF srgba(0,160,192,1)
167: (128,160,192,255) #80A0C0FF srgba(128,160,192,1)
168: ( 64, 32, 64,255) #402040FF srgba(64,32,64,1)
169: (192, 32, 64,255) #C02040FF srgba(192,32,64,1)
170: ( 64,160, 64,255) #40A040FF srgba(64,160,64,1)
171: (192,160, 64,255) #C0A040FF srgba(192,160,64,1)
172: ( 64, 32,192,255) #4020C0FF srgba(64,32,192,1)
173: (192, 32,192,255) #C020C0FF srgba(192,32,192,1)
174: ( 64,160,192,255) #40A0C0FF srgba(64,160,192,1)
175: (192,160,192,255) #C0A0C0FF srgba(192,160,192,1)
176: ( 0, 96, 64,255) #006040FF srgba(0,96,64,1)
177: (128, 96, 64,255) #806040FF srgba(128,96,64,1)
178: ( 0,224, 64,255) #00E040FF srgba(0,224,64,1)
179: (128,224, 64,255) #80E040FF srgba(128,224,64,1)
180: ( 0, 96,192,255) #0060C0FF srgba(0,96,192,1)
181: (128, 96,192,255) #8060C0FF srgba(128,96,192,1)
182: ( 0,224,192,255) #00E0C0FF srgba(0,224,192,1)
183: (128,224,192,255) #80E0C0FF srgba(128,224,192,1)
184: ( 64, 96, 64,255) #406040FF srgba(64,96,64,1)
185: (192, 96, 64,255) #C06040FF srgba(192,96,64,1)
186: ( 64,224, 64,255) #40E040FF srgba(64,224,64,1)
187: (192,224, 64,255) #C0E040FF srgba(192,224,64,1)
188: ( 64, 96,192,255) #4060C0FF srgba(64,96,192,1)
189: (192, 96,192,255) #C060C0FF srgba(192,96,192,1)
190: ( 64,224,192,255) #40E0C0FF srgba(64,224,192,1)
191: (192,224,192,255) #C0E0C0FF srgba(192,224,192,1)
192: ( 32, 32, 0,255) #202000FF srgba(32,32,0,1)
193: (160, 32, 0,255) #A02000FF srgba(160,32,0,1)
194: ( 32,160, 0,255) #20A000FF srgba(32,160,0,1)
195: (160,160, 0,255) #A0A000FF srgba(160,160,0,1)
196: ( 32, 32,128,255) #202080FF srgba(32,32,128,1)
197: (160, 32,128,255) #A02080FF srgba(160,32,128,1)
198: ( 32,160,128,255) #20A080FF srgba(32,160,128,1)
199: (160,160,128,255) #A0A080FF srgba(160,160,128,1)
200: ( 96, 32, 0,255) #602000FF srgba(96,32,0,1)
201: (224, 32, 0,255) #E02000FF srgba(224,32,0,1)
202: ( 96,160, 0,255) #60A000FF srgba(96,160,0,1)
203: (224,160, 0,255) #E0A000FF srgba(224,160,0,1)
204: ( 96, 32,128,255) #602080FF srgba(96,32,128,1)
205: (224, 32,128,255) #E02080FF srgba(224,32,128,1)
206: ( 96,160,128,255) #60A080FF srgba(96,160,128,1)
207: (224,160,128,255) #E0A080FF srgba(224,160,128,1)
208: ( 32, 96, 0,255) #206000FF srgba(32,96,0,1)
209: (160, 96, 0,255) #A06000FF srgba(160,96,0,1)
210: ( 32,224, 0,255) #20E000FF srgba(32,224,0,1)
211: (160,224, 0,255) #A0E000FF srgba(160,224,0,1)
212: ( 32, 96,128,255) #206080FF srgba(32,96,128,1)
213: (160, 96,128,255) #A06080FF srgba(160,96,128,1)
214: ( 32,224,128,255) #20E080FF srgba(32,224,128,1)
215: (160,224,128,255) #A0E080FF srgba(160,224,128,1)
216: ( 96, 96, 0,255) #606000FF srgba(96,96,0,1)
217: (224, 96, 0,255) #E06000FF srgba(224,96,0,1)
218: ( 96,224, 0,255) #60E000FF srgba(96,224,0,1)
219: (224,224, 0,255) #E0E000FF srgba(224,224,0,1)
220: ( 96, 96,128,255) #606080FF srgba(96,96,128,1)
221: (224, 96,128,255) #E06080FF srgba(224,96,128,1)
222: ( 96,224,128,255) #60E080FF srgba(96,224,128,1)
223: (224,224,128,255) #E0E080FF srgba(224,224,128,1)
224: ( 32, 32, 64,255) #202040FF srgba(32,32,64,1)
225: (160, 32, 64,255) #A02040FF srgba(160,32,64,1)
226: ( 32,160, 64,255) #20A040FF srgba(32,160,64,1)
227: (160,160, 64,255) #A0A040FF srgba(160,160,64,1)
228: ( 32, 32,192,255) #2020C0FF srgba(32,32,192,1)
229: (160, 32,192,255) #A020C0FF srgba(160,32,192,1)
230: ( 32,160,192,255) #20A0C0FF srgba(32,160,192,1)
231: (160,160,192,255) #A0A0C0FF srgba(160,160,192,1)
232: ( 96, 32, 64,255) #602040FF srgba(96,32,64,1)
233: (224, 32, 64,255) #E02040FF srgba(224,32,64,1)
234: ( 96,160, 64,255) #60A040FF srgba(96,160,64,1)
235: (224,160, 64,255) #E0A040FF srgba(224,160,64,1)
236: ( 96, 32,192,255) #6020C0FF srgba(96,32,192,1)
237: (224, 32,192,255) #E020C0FF srgba(224,32,192,1)
238: ( 96,160,192,255) #60A0C0FF srgba(96,160,192,1)
239: (224,160,192,255) #E0A0C0FF srgba(224,160,192,1)
240: ( 32, 96, 64,255) #206040FF srgba(32,96,64,1)
241: (160, 96, 64,255) #A06040FF srgba(160,96,64,1)
242: ( 32,224, 64,255) #20E040FF srgba(32,224,64,1)
243: (160,224, 64,255) #A0E040FF srgba(160,224,64,1)
244: ( 32, 96,192,255) #2060C0FF srgba(32,96,192,1)
245: (160, 96,192,255) #A060C0FF srgba(160,96,192,1)
246: ( 32,224,192,255) #20E0C0FF srgba(32,224,192,1)
247: (160,224,192,255) #A0E0C0FF srgba(160,224,192,1)
248: ( 96, 96, 64,255) #606040FF srgba(96,96,64,1)
249: (224, 96, 64,255) #E06040FF srgba(224,96,64,1)
250: ( 96,224, 64,255) #60E040FF srgba(96,224,64,1)
251: (224,224, 64,255) #E0E040FF srgba(224,224,64,1)
252: ( 96, 96,192,255) #6060C0FF srgba(96,96,192,1)
253: (224, 96,192,255) #E060C0FF srgba(224,96,192,1)
254: ( 96,224,192,255) #60E0C0FF srgba(96,224,192,1)
255: (224,224,192,255) #E0E0C0FF srgba(224,224,192,1)
Rendering intent: Perceptual
Gamma: 0.454545
Chromaticity:
red primary: (0.64,0.33)
green primary: (0.3,0.6)
blue primary: (0.15,0.06)
white point: (0.3127,0.329)
Matte color: grey74
Background color: white
Border color: srgb(223,223,223)
Transparent color: none
Interlace: None
Intensity: Undefined
Compose: Over
Page geometry: 625x468+0+0
Dispose: Undefined
Iterations: 0
Compression: Zip
Orientation: Undefined
Properties:
date:create: 2020-01-21T11:05:35+00:00
date:modify: 2020-01-21T11:05:35+00:00
png:IHDR.bit-depth-orig: 8
png:IHDR.bit_depth: 8
png:IHDR.color-type-orig: 3
png:IHDR.color_type: 3 (Indexed)
png:IHDR.interlace_method: 0 (Not interlaced)
png:IHDR.width,height: 625, 468
png:PLTE.number_colors: 256
png:sRGB: intent=0 (Perceptual Intent)
signature: 12c7d4ddd28f23d431ea875c618757f305577bf8aec476ec418728c161d3de8e
Artifacts:
verbose: true
Tainted: False
Filesize: 6497B
Number pixels: 292500
Pixels per second: 54.4078MP
User time: 0.000u
Elapsed time: 0:01.005
Version: ImageMagick 7.0.9-6 Q16 x86_64 2019-11-27 https://imagemagick.org
If ImageMagick is too big a dependency for you, you can also use the much smaller pngcheck like this:如果ImageMagick对你来说依赖太大，你也可以像这样使用小得多的pngcheck ：
pngcheck -vp YourImage.png
Sample Output样本输出
File: YourImage.png (6497 bytes)
chunk IHDR at offset 0x0000c, length 13
625 x 468 image, 8-bit palette, non-interlaced
chunk PLTE at offset 0x00025, length 768: 256 palette entries
0: ( 0, 0, 0) = (0x00,0x00,0x00)
1: (128, 0, 0) = (0x80,0x00,0x00)
2: ( 0,128, 0) = (0x00,0x80,0x00)
3: (128,128, 0) = (0x80,0x80,0x00)
4: ( 0, 0,128) = (0x00,0x00,0x80)
5: (128, 0,128) = (0x80,0x00,0x80)
6: ( 0,128,128) = (0x00,0x80,0x80)
7: (128,128,128) = (0x80,0x80,0x80)
8: ( 64, 0, 0) = (0x40,0x00,0x00)
9: (192, 0, 0) = (0xc0,0x00,0x00)
10: ( 64,128, 0) = (0x40,0x80,0x00)
11: (192,128, 0) = (0xc0,0x80,0x00)
12: ( 64, 0,128) = (0x40,0x00,0x80)
13: (192, 0,128) = (0xc0,0x00,0x80)
14: ( 64,128,128) = (0x40,0x80,0x80)
15: (192,128,128) = (0xc0,0x80,0x80)
16: ( 0, 64, 0) = (0x00,0x40,0x00)
17: (128, 64, 0) = (0x80,0x40,0x00)
18: ( 0,192, 0) = (0x00,0xc0,0x00)
19: (128,192, 0) = (0x80,0xc0,0x00)
20: ( 0, 64,128) = (0x00,0x40,0x80)
21: (128, 64,128) = (0x80,0x40,0x80)
22: ( 0,192,128) = (0x00,0xc0,0x80)
23: (128,192,128) = (0x80,0xc0,0x80)
24: ( 64, 64, 0) = (0x40,0x40,0x00)
25: (192, 64, 0) = (0xc0,0x40,0x00)
...
...
251: (224,224, 64) = (0xe0,0xe0,0x40)
252: ( 96, 96,192) = (0x60,0x60,0xc0)
253: (224, 96,192) = (0xe0,0x60,0xc0)
254: ( 96,224,192) = (0x60,0xe0,0xc0)
255: (224,224,192) = (0xe0,0xe0,0xc0)
chunk IDAT at offset 0x00331, length 5660
zlib: deflated, 32K window, default compression
chunk IEND at offset 0x01959, length 0
No errors detected in p.png (4 chunks, 97.8% compression).
I see you want to write your results out with the same palette as you started with.我看到您想用与开始时相同的调色板写出结果。 I don't believe OpenCV can do that, but you could write your results as a normal, unpalettised PNG with OpenCV and then remap that to your original palette with ImageMagick in Terminal.我不相信 OpenCV 可以做到这一点，但是您可以使用 OpenCV 将结果编写为普通的、未调色的 PNG，然后在终端中使用ImageMagick将其重新映射到您的原始调色板。
So, we can make a 7x1 pixel image of the colours in your palette like this:因此，我们可以像这样制作调色板中颜色的 7x1 像素图像：
magick -size 1x1 xc:black xc:navy xc:green xc:'#8080C0' xc:'#C00080' xc:'#C04000' xc:'#E0E0C0' +append palette.png
That looks like this when enlarged:放大后看起来是这样的：
You can now make any results images you like with OpenCV, but convert (ie "remap" ) them to that original palette like this:您现在可以使用 OpenCV 制作您喜欢的任何结果图像，但将它们转换（即“重新映射” ）为原始调色板，如下所示：
magick OpenCVresult.png +dither -remap palette.png result.png
If your ImageMagick is v6, use convert in place of magick .如果您的 ImageMagick 是 v6，请使用convert代替magick 。
声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.
相关问题
在OpenCV Python中使用1通道图像作为遮罩 - Using a 1-channel image as a mask in opencv python
Python Pillow：如何从 1 通道图像生成 3 通道图像？ - Python Pillow: how to produce 3-channel image from 1-channel image?
如何使用opencv为python编写灰色（1通道）图像 - how to write gray (1-channel) image with opencv for python
使用模式P将阵列转换为1通道图像 - Convert Array to 1-Channel Image with mode P
灰度图像的 OpenCV 读取返回 3 通道（颜色？）数组而不是单通道 - OpenCV imread of grayscale image returns 3-channel (color?) array instead of single-channel
如何将1通道numpy矩阵转换为4通道单色图像 - How to convert 1-channel numpy matrix to 4-channel monochromatic image
使用 PIL 将 alpha 通道添加到图像 - add alpha channel to an image with PIL
使用最近邻插值调整 1 通道 numpy（图像）数组的大小 - Resize 1-channel numpy (image) array with nearest neighbour interpolation
如何使用PIL将1通道图像转换为3通道？ - How to convert a 1 channel image into a 3 channel with PIL?
OpenCV Python：3通道float32图像读取的快速解决方案？ - OpenCV Python: fast solution for 3-channel float32 image reading?
相关标签