<?xml version="1.0" encoding="utf-8" ?><rss version="2.0"><channel><title><![CDATA[GAN_player的博客]]></title><description><![CDATA[]]></description><link>https://blog.youkuaiyun.com/GAN_player</link><language>zh-cn</language><generator>https://blog.youkuaiyun.com/</generator><copyright><![CDATA[Copyright &copy; GAN_player]]></copyright><item><title><![CDATA[Markdown编辑神器-Typora]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/96755051</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/96755051</guid><author>GAN_player</author><pubDate>Sun, 21 Jul 2019 20:03:45 +0800</pubDate><description><![CDATA[Markdown For Typora
Overview
Markdown is created by Daring Fireball, the original guideline is here. Its syntax, however, varies between different parsers or editors. Typora is using [GitHub Flavored ...]]></description><category></category></item><item><title><![CDATA[如何计算fpr95]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/85113431</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/85113431</guid><author>GAN_player</author><pubDate>Thu, 20 Dec 2018 10:44:46 +0800</pubDate><description><![CDATA[对于二分类问题，我们经常通过ROC曲线及FPR95来判断分类器的好坏。这里提供两种方法。

一种是sklearn.metrics中的roc_curve包，可直接用于计算在不同阈值下，TPR和FPR对应的值，进而可以得出TPR=0.95时，FPR的值。

&quot;&quot;&quot;
label=1表示正样本，scores为预测概率，数值越大，越有可能是正样本
&quot;&quot;&quot;
from sklearn.metrics impor...]]></description><category></category></item><item><title><![CDATA[tmux常见指令合集]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/84938224</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/84938224</guid><author>GAN_player</author><pubDate>Mon, 10 Dec 2018 13:02:32 +0800</pubDate><description><![CDATA[在这里插入代码片
# 新增
$ tmux
# OR
$ tmux new -s &amp;lt;your_session_name&amp;gt;

# session 列表
$ tmux ls

# 重新連線 session
$ tmux a -t 0
# OR
$ tmux a -t &amp;lt;session_name&amp;gt;

# 刪除 session
$ tmux kill-session -t 0
# O...]]></description><category></category></item><item><title><![CDATA[如何彻底打乱一个数组]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/83863987</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/83863987</guid><author>GAN_player</author><pubDate>Thu, 08 Nov 2018 16:49:16 +0800</pubDate><description><![CDATA[import numpy as np

def abs_shuffle(arr0):
  arr = arr0.copy()
  for i in np.arange(len(arr)-1,0,-1):
    idx = np.random.choice(range(i))
    arr[i], arr[idx] = arr[idx], arr[i]
  return arr


arr1 =...]]></description><category></category></item><item><title><![CDATA[python实现图像直方图均衡]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/81512199</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/81512199</guid><author>GAN_player</author><pubDate>Wed, 08 Aug 2018 17:26:22 +0800</pubDate><description><![CDATA[def histeq(img,nbr_bins=256):
    &quot;&quot;&quot; Histogram equalization of a grayscale image. &quot;&quot;&quot;


    imhist, bins = np.histogram(img.flatten(), nbr_bins, normed = True)

    cdf = imhist.cumsum() # cumulative...]]></description><category></category></item><item><title><![CDATA[python实现投影变换]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/81512183</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/81512183</guid><author>GAN_player</author><pubDate>Wed, 08 Aug 2018 17:25:17 +0800</pubDate><description><![CDATA[def warp_perspective(src, M, width, height, origin_x, origin_y, flags=cv2.INTER_NEAREST, borderMode=cv2.BORDER_CONSTANT,
                     borderValue=0, dst=None):
    &amp;quot;&amp;quot;&amp;quot;
    Implementation in Py...]]></description><category></category></item><item><title><![CDATA[tf.nn.conv1d]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/79373480</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/79373480</guid><author>GAN_player</author><pubDate>Mon, 26 Feb 2018 09:40:32 +0800</pubDate><description><![CDATA[
  inputs=tf.ones((64,64,3)) 
  w=tf.constant(1,tf.float32,(5,3,32)) 
  tf.nn.conv1d(inputs,w,2,’SAME’) 
  Out[38]: ]]></description><category></category></item><item><title><![CDATA[tensorboard使用问题]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78858736</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78858736</guid><author>GAN_player</author><pubDate>Wed, 20 Dec 2017 23:17:28 +0800</pubDate><description><![CDATA[1。如果http://localhost:6006/ 进不了，可以换成这个网址： http://127.0.0.1:6006/ 
2。如果出现No scalar data was found…  
先检查代码有没有问题，如果没有问题就是tensorboard指令用错了。应该在events文件的上一层路径输入：tensorboard –logdir=summary 
注意，等号两边，不能有空格，不能有]]></description><category></category></item><item><title><![CDATA[ubuntu下如何在spyder上输入汉字]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78850433</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78850433</guid><author>GAN_player</author><pubDate>Wed, 20 Dec 2017 10:35:41 +0800</pubDate><description><![CDATA[每次写python脚本都是在spyder编辑器上写，然而在ubuntu环境中默认无法输入汉字。百度找了很久，没有找到解决方法，突然有一天碰巧成功，遂记录下来，以免日后遗忘。 
１．进入system 
２. 进入language support 
 
 3.选择regional formats 
  
 4.点击apply system-wide　 
 就ＯＫ啦]]></description><category></category></item><item><title><![CDATA[如何在英文环境下安装搜狗输入法]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78678360</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78678360</guid><author>GAN_player</author><pubDate>Thu, 30 Nov 2017 17:07:21 +0800</pubDate><description><![CDATA[1、官网下载搜狗.deb安装包
2、sudo apt-get install fcitx
3、sudo dpkg -i sougoupinyinXXX.deb #运行完这部会出错，显示缺很多东西，不用管，直接进行第四步
4、sudo apt-get -f install
5、再次执行第三步：sudo dpkg -i sougoupinyinXXX.deb
安装完成之后，重启电脑6、系统设置>语言支持]]></description><category></category></item><item><title><![CDATA[matplotlib调整子图间距，调整整体空白]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78543643</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78543643</guid><author>GAN_player</author><pubDate>Wed, 15 Nov 2017 18:17:18 +0800</pubDate><description><![CDATA[fig.tight_layout()#调整整体空白
  plt.subplots_adjust(wspace =0, hspace =0)#调整子图间距]]></description><category></category></item><item><title><![CDATA[matplotlib如何不显示图片只保存图片]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78543288</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78543288</guid><author>GAN_player</author><pubDate>Wed, 15 Nov 2017 17:42:39 +0800</pubDate><description><![CDATA[matplotlib迭代次数多了之后，如果每次都显示，会卡顿，因此下面这种方法可以不显示，直接保存import numpy as np
import matplotlib
matplotlib.use('Agg')
from matplotlib.pyplot import plot,savefig
x=np.linspace(-4,4,30)
y=np.sin(x);
plot(x,y,'--*b]]></description><category></category></item><item><title><![CDATA[opencv——利用轮廓信息画斜矩形、外接圆、拟合直线等]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78368947</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78368947</guid><author>GAN_player</author><pubDate>Fri, 27 Oct 2017 20:44:56 +0800</pubDate><description><![CDATA[#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Oct 27 20:14:10 2017@author: fs
"""import cv2
import numpy as np
img = cv2.imread('lunkuo1.png')ret,thresh = cv2.threshold(cv2.cvtColor]]></description><category></category></item><item><title><![CDATA[opencv——计算轮廓的周长、所包含面积、中心点]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78368687</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78368687</guid><author>GAN_player</author><pubDate>Fri, 27 Oct 2017 20:06:23 +0800</pubDate><description><![CDATA[#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Oct 27 19:14:12 2017@author: fs
"""import cv2
import numpy as np
img = cv2.imread('lunkuo.png')ret,thresh = cv2.threshold(cv2.cvtColor(]]></description><category></category></item><item><title><![CDATA[opencv——利用图像金字塔合成图片]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78359154</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78359154</guid><author>GAN_player</author><pubDate>Thu, 26 Oct 2017 21:48:29 +0800</pubDate><description><![CDATA[import cv2
import numpy as np
DEPTH =3imgA = cv2.imread('apple.jpg')
imgB = cv2.imread('orange.jpg')
gA=[imgA]
lpA=[]
gB=[imgB]
lpB=[]
for i in range(DEPTH):
  tempA1 = cv2.pyrDown(gA[i])
  tempB1 = cv]]></description><category></category></item><item><title><![CDATA[opencv——绘制图像轮廓]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78359142</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78359142</guid><author>GAN_player</author><pubDate>Thu, 26 Oct 2017 21:46:53 +0800</pubDate><description><![CDATA["""
Created on Thu Oct 26 21:40:07 2017@author: fs
绘制轮廓
函数 cv2.findContours() 有三个参数,第一个是输入图像,第二个是
轮廓检索模式,第三个是轮廓近似方法。返回值有三个,第一个是图像,第二个
是轮廓,第三个是(轮廓的)层析结构。轮廓(第二个返回值)是一个 Python
列表,其中存储这图像中的所有轮廓。每一个轮廓都是一个 N]]></description><category></category></item><item><title><![CDATA[opencv——canny算子提取图像边缘]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78346633</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78346633</guid><author>GAN_player</author><pubDate>Wed, 25 Oct 2017 20:59:25 +0800</pubDate><description><![CDATA[import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('test.png',0)
edges = cv2.Canny(img,150,200)
plt.subplot(121),plt.imshow(img,cmap = 'gray')
plt.title('Original Image]]></description><category></category></item><item><title><![CDATA[opencv的二值化1（ 用Otsu’s二值化，找到双峰阈值）]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78346596</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78346596</guid><author>GAN_player</author><pubDate>Wed, 25 Oct 2017 20:57:05 +0800</pubDate><description><![CDATA[import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('thresh1.PNG',0)
# global thresholding
ret1,th1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
# Otsu's thresholding]]></description><category></category></item><item><title><![CDATA[opencv图像开运算闭运算/腐蚀膨胀]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78346577</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78346577</guid><author>GAN_player</author><pubDate>Wed, 25 Oct 2017 20:55:20 +0800</pubDate><description><![CDATA['''
腐蚀：根据卷积核的大小靠近前景的所有像素都会被腐蚀
掉(变为 0),所以前景物体会变小,整幅图像的白色区域会减少。这对于去除
白噪声很有用,也可以用来断开两个连在一块的物体等。
总结：腐蚀会减少白色物体的面积
膨胀: 与腐蚀相反,与卷积核对应的原图像的像素值中只要有一个是 1,中心元
素的像素值就是 1。所以这个操作会增加图像中的白色区域(前景)。一般在去
噪声时先用腐蚀再用膨胀。因为腐蚀在]]></description><category></category></item><item><title><![CDATA[opencv各种滤波器的用法]]></title><link>https://blog.youkuaiyun.com/GAN_player/article/details/78343850</link><guid>https://blog.youkuaiyun.com/GAN_player/article/details/78343850</guid><author>GAN_player</author><pubDate>Wed, 25 Oct 2017 17:17:22 +0800</pubDate><description><![CDATA[import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('test2.jpg')
kernel = np.ones((5,5),np.float32)/25
dst = cv2.filter2D(img,-1,kernel)#blur = cv2.GaussianBlur(img,(5,5]]></description><category></category></item></channel></rss>