二、开始使用视频.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2>开始使用视频</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>目标</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<ul>\n",
    "    <li>学会读取视频、显示视频和保存视频</li>\n",
    "    <li>学会从摄像头中捕捉并显示它</li>\n",
    "    <li>你将要学习这两个函数：cv2.VideoCapture(), cv2.VideoWriter()</li>\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>2.1 从摄像头中捕捉视频</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>通常，我们必须用摄像头捕捉实时图像。OpenCv 为此提供了一个非常简单的接口。让我们使用摄像头捕获一段视频（我正使用我的笔记本电脑内置的摄像头），将它转换成灰度视频并显示它。这是是一个简单的开始。</P>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>为了捕获视频，你需要创建一个 VideoCapture 对象。它的参数可以是设备索引，也可以是视频文件的名称。设备索引就是指定的摄像头的编号。一般的笔记本中都有摄像头（比如在我的例子中）。所以我只传递 0 （或者 -1）。你可以通过传递 1 选择第二个摄像头等。之后，你就可以一帧一帧的捕获视频里。但是在最后不要忘了停止捕获视频。</P>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "\n",
    "cap = cv2.VideoCapture(0)\n",
    "\n",
    "while(True):\n",
    "    #一帧一帧捕获视频 \n",
    "    #ret的值为True 或 False 。frame 表示当前截取一帧的图片\n",
    "    ret, frame = cap.read()\n",
    "    \n",
    "    #在这里，我们在每一帧上操作\n",
    "    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)\n",
    "    \n",
    "    #显示帧的结果\n",
    "    cv2.imshow('frame',gray)\n",
    "    if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "        break\n",
    "        \n",
    "#当一切结束时，释放资源\n",
    "cap.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>cap.read() 返回一个布尔值（True ／ False）。如果帧读取是正确的，它会返回True。因此你可以通过检查它的返回值确定视频是否到达了文件尾。\n",
    "    </p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>有时，cap 不能正确的初始化捕获。这种情况下，上面的代码会报错。你可以检查它是否进行了初始化通过 cap.isOpened() 方法。如果返回值\n",
    "是True，那么则没问题。否则，就使用 cap.open()。</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>你可以使用函数 cap.get(prold) 来获得视频的一些参数信息。这里propld 可以是1-18的任意整数。每一个数字表示视频的一个属性（如果它适用于该视频），可以在这里看到完整的详细信息：<a href=\"http://docs.opencv.org/modules/highgui/doc/reading_and_writing_images_and_video.html#videocapture-get\">属性标识符</a>，其中的一些值可以用 cap.set(propld,value) 进行修改。value 是你想要的新值。</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<p>例如，我可以使用 cap.get(3) 和 cap.get(4) 来查看每一帧的宽和高。默认情况下得到的值是 640x480 。但是我们可以使用 ret=cap.set(3,320) 和 ret=cap.set(4,240) 来把宽和高改成320x240</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h4>注意</h4>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当你的程序报错时，你首先应该检查的是你的摄像头能否在其他的程序中正常工作（比如Linux 下的 Cheese）"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>2.2 从文件中播放视频</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "与从摄像头中捕获相同，只要将摄像头的设备索引号改成视频的文件名称即可。在播放每一帧时，使用cv2.waitKey() 设置适当的持续时间。如果设置的太低视频就会播放的非常快，如果设置的太高，视频就会播放的特别慢（当然，这就是你如何使用慢镜头显示视频的方法）。25毫秒在正常情况下是可以的。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "#改写了原文档中代码 改写后代码在视频播放完后，会自动退出循环\n",
    "cap = cv2.VideoCapture('video1.avi')\n",
    "\n",
    "while(cap.isOpened()):\n",
    "    ret, frame = cap.read()\n",
    "    if ret == True:\n",
    "        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)\n",
    "\n",
    "        cv2.imshow('frame',gray)\n",
    "        if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "            break\n",
    "    else :\n",
    "        break\n",
    "cap.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h4>注意</h4>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "你应该确保你已经正确安装了合适版本的 ffmpeg 或者 gstreamer。如果你装错了那就比较头疼了。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>2.3 保存视频</h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "我们捕获一个视频，然后一帧帧处理并且想保存这个视频。对于图像来说它是非常简单的，只需要使用 cv.imwrite().但是在这里，需要做更多的工作。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "这次我们创建一个 VideoWriter 对象。我们应该指定一个输出文件名称（比如：out.avi) 然后，我们需要指定 FourCC 编码（下面将要介绍）。每秒的帧数和帧大小也需要确定。最后一个时 isColor 标志，如果时 True\n",
    "，每一帧就是彩色图，否则就是灰色图。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "FourCC 是一个四字节码，用来确定视频的编码格式。可用的编码列表可以从 fourcc.org 查到。这是依赖于平台，下面的编码器对我来说没问题。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<ul>\n",
    "    <li>In Fedora: DIVX, XVID, MJPG, X264, WMV1, WMV2. (XVID 更加适合. MJPG 的结果是高尺寸的视频. X264 提供非常小的视频)</li>\n",
    "    <li>In Windows: DIVX (有待测试和添加)</li>\n",
    "    <li>In OSX :(我没有OSX的权限。有人能够填这张表吗？)</li>\n",
    "    </ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "FourCC 码是以下面的格式传给程序，以MJPG为例：cv2.VideoWriter_fourcc('M','J','P','G') 或者 cv2.VideoWriter_fourcc(*'MJPG)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "下面的代码是从摄像投中捕获视频，沿水平方向旋转每一帧并保存它。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import cv2\n",
    "\n",
    "cap = cv2.VideoCapture(0)\n",
    "\n",
    "# 定义编码格式并且创建 VideoWriter 对象\n",
    "fourcc = cv2.VideoWriter_fourcc(*'XVID')\n",
    "out = cv2.VideoWriter('output.avi',fourcc, 20.0, (640,480))\n",
    "\n",
    "while(cap.isOpened()):\n",
    "    ret, frame = cap.read()\n",
    "    if ret==True:\n",
    "        frame = cv2.flip(frame,0)\n",
    "\n",
    "        # write the flipped frame\n",
    "        out.write(frame)\n",
    "\n",
    "        cv2.imshow('frame',frame)\n",
    "        if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "            break\n",
    "    else:\n",
    "        break\n",
    "\n",
    "# 如果工作结束则释放掉所有资源\n",
    "cap.release()\n",
    "out.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "下面代码为：MAC 版本"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# coding:utf-8\n",
    "import cv2\n",
    "import sys\n",
    "\n",
    "\n",
    "cap = cv2.VideoCapture(0)\n",
    "cap.set(3,640)\n",
    "cap.set(4,480)\n",
    "cap.set(1, 10.0)\n",
    "#此处fourcc的在MAC上有效，如果视频保存为空，那么可以改一下这个参数试试, 也可以是-1\n",
    "fourcc = cv2.VideoWriter_fourcc('m', 'p', '4', 'v')\n",
    "# 第三个参数则是镜头快慢的，10为正常，小于10为慢镜头\n",
    "out = cv2.VideoWriter('output2.avi', fourcc,10,(640,480))\n",
    "while True:\n",
    "    ret,frame = cap.read()\n",
    "    if ret == True:\n",
    "        frame = cv2.flip(frame, 1)\n",
    "        a = out.write(frame)\n",
    "        cv2.imshow(\"frame\", frame)\n",
    "        if cv2.waitKey(1) & 0xFF == ord('q'):\n",
    "            break\n",
    "    else:\n",
    "        break\n",
    "cap.release()\n",
    "out.release()\n",
    "cv2.destroyAllWindows()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}