[Python] OpenCV matchTemplate

http://zwindr.blogspot.tw/2017/02/python-opencv-matchtemplate.html

OpenCV中感兴趣区域的选取与检测

http://blog.csdn.net/qq_23968185/article/details/51429732

Python 3 教程

http://www.runoob.com/python3/python3-tutorial.html

http://glophy.com/index.php/2014-01-26-07-09-30

Python去除列表list重複或相似元素的方法

http://www.sijitao.net/2441.html

def remove_similar ( lists , similarity = 0.9 ): 
    i = 0 
    l = len ( lists ) while i < l : 
        j = i + 1 while j < l : 
            seq = difflib . SequenceMatcher ( None , lists [ i ], lists [ j ]) 
            ratio =
    
        seq . ratio () if ratio >= similarity : del lists [ j ] 
                l = l - 1 else : 
                j += 1 
        i += 1 
        return lists

OpenCV程序在Debug時出現「PDB文件無法加載」

VS的 [Tools]->[Options]->[Debugging]->[General]

勾選Enable source server support 選項

然後 [Tools]->[Options]->[Debugging]->[Symbols]

勾選那個Microsoft Symbol Server，然後確定

原文網址：https://read01.com/xmxRx.html

Tensorflow 1.4 gpu版 安裝步驟

0.安裝Pycharm 2017.1.5

1.安裝cuda_8.0.61_win10.exe

2.下載cudnn-8.0-windows10-x64-v6.0,
將bin, lib, include裡面檔案放到C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\xxx

3.安裝Anaconda3-4.1.1-Windows-x86_64.exe
安裝到C:\Anaconda3

4.Anaconda Prompt (or CMD)
conda create -n tensorflow-gpu python=3.5.2
activate tensorflow-gpu (activate後的動作都會安裝到C:\Anaconda3\envs\tensorflow-gpu裡面的python套件管理)
pip install tensorflow-gpu keras pillow h5py (如果有缺套件一樣從這邊新增, 可用pip list檢查)

5.pycharm new project
Interpreter : C:\Anaconda3\envs\tensorflow-gpu\python.exe (應該會需要一點時間，pycharm右下角 Background Task)





# 參考網址 http://blog.nitishmutha.com/tensorflow/2017/01/22/TensorFlow-with-gpu-for-windows.html

# 版本需對應 cuda8 + cuDNN v6

Tensorflow 1.5rc1 build on windows10

Tensorflow 1.4 只能用cuda 8
visual studio 2017 只能搭配cuda 9
cuda 8 需要用到 visual studio 2015的runtime
visual studio 2015的runtime跟visual studio 2017無法共存

因此 Tensorflow 1.4 跟 visual studio 2017 無法共存...

所以預計使用Tensorflow 1.5rc1, 但因為是prebuild, 沒有windows版本的

所以要自行從source code安裝

https://github.com/tensorflow/tensorflow/tree/r0.12/tensorflow/contrib/cmake

Pre-requisites

• CMake version 3.5 up to 3.6
• Git
• SWIG
• Additional pre-requisites for Microsoft Windows:
  • Visual Studio 2015
  • Python 3.5
  • NumPy 1.11.0 or later
• Additional pre-requisites for Linux:
  • Python 2.7 or later
  • Docker (for automated testing)
  • NumPy 1.11.0 or later

download CMake, Git, SWIG, cuDNN 9.1, cuda9

.... 2018/1/23 編譯失敗，版本匹配問題仍多，暫時放棄等待新版本tensorflow套件