如何从numpy数组的概率图中获取二进制图像？
[英]How can I get a binary image from a probability map in a numpy array?
问题描述
Hi I have the following heat map (probability map) of an object. 
嗨，我有一个物体的以下热图（概率图）。 How can I convert this heatmap to a binary image?
如何将此热图转换为二进制图像？ I want to take the maximum area of the object to increase the intersection with the image in the right side. 我想获取对象的最大面积，以增加与右侧图像的交集。 I really appreciate any help. 我非常感谢您的帮助。
1 个解决方案
解决方案1
0 2017-06-10 01:02:04
To binarize, compare against a threshold: 要进行二值化，请与阈值进行比较：
thresh = 0.5
bin_map = prob_map > thresh
From there you can use a very large variety of solutions: 从那里，您可以使用各种各样的解决方案：
Saving a numpy array as an image 将numpy数组另存为图像
I recommend this one from migas: 我从migas推荐这个：
from PIL import Image
im = Image.fromarray(A)
im.save("your_file.jpeg")
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.