镜像驱动截屏技术在数字语音教室中的应用

镜像驱动截屏技术在数字语音教室中的应用

论文编号:TX256 论文字数:11313,页数:42 有开题报告，任务书

摘要
数字语音教学、远程教学等网络应用要求实时高效快捷地捕获PC系统上的屏幕信息。采用传统的API截屏方式，通过调用windows系统拷贝函数来截取屏幕信息的方法会耗费大量的CPU资源, 本文采用镜像驱动程序，在只耗费5%—10%的CPU资源情况下每秒截取10至20帧屏幕图像，达到实时捕获PC系统屏幕信息的目的。
这些新型业务的另一特点是，服务端会有流媒体视频文件播放，接收端要能有连贯的视觉效果。在这种方式下，用传统API截屏方式就会有以下几个缺陷：1. 可能捕获不到流媒体视频画面，这是因为视频文件在播放过程中系统往往调用DirectDraw来加速视频的播放，由于DirectDraw的特殊性，直接采用API函数来截屏的话，会捕获不到视频画面信息。本文采用镜像驱动，在设备层中就能捕获到屏幕信息，Mirror Drive的好处就是可以直接利用硬件，达到用户模式下让人难以企及的“实时”，这是因为使用BitBlt拷贝屏幕内容，其数据需要从显存传输到系统内存，经过低速繁忙的系统总线，自然比从内存到内存的数据拷贝慢得多。2. 严重浪费网络资源，一张1024*768屏幕位图约为4M，每秒20帧就是80M，如此大的数据量在一些带宽较低的链路上，极可能引起严重的通信瓶颈。本文采用JPEG图像压缩算法在视觉上不损图像质量的情况下将捕获的位图高效的压缩，再将压缩后的图像传输给客户端，即使有成千上万的客户端，也不会造成网络崩溃。

关键字：镜像驱动；位图；API；DirectDraw；JPEG图像压缩算法

The Application of Mirror-driven Screen Capture Technology in Digitized Voice Classrooms

Abstract
Digitized Voice classrooms, distance learning and other network applications require real-time high speed to catch on the PC screen information. Traditional API methods, By calling the Windows system function to copy the interception of the screen information will consume lots of CPU resources, In this paper, mirror driver, only 5% -10% of the cost of CPU resources per second interception of 10-20 screen images, achieve real-time PC screen capture information purposes.
Another feature of these new businesses is the server will be Video Streaming Media Player, receiver must have a coherent visual effect. In this way, using traditional API methods will have the following shortcomings: 1. Can not catch the streaming media video images. This is because video broadcast in the process of systems are often called DirectDraw to accelerate video broadcast Due to the special nature of DirectDraw, the API used directly to the province, Video capture will not screen information. In this paper, mirror-driven, the equipment will be able to capture layer of the screen information, Mirror Driver advantage is the direct use of hardware, user mode to make the enterprises and the "real-time", it is because the use BitBlt copy screen contents Its data transmission needs from memory to memory, after low-speed busiest bus, natural than from memory to memory to copy data much slower. 2. A serious waste of network resources, a 1024 * 768 pixel screen bitmap is about 4M, 20 per second is 80M. Such a large volume of data transport in some low bandwidth link, is very likely to lead to a serious bottleneck in communications. In this paper, using the JPEG image compression algorithm in the visual without compromising the quality of images will be captured under the bitmap efficient compression, and then compressed image transmission to the client, even if there are thousands of clients, will not cause the collapse of the network.

Keywords: Mirror drive; Bitmap; API; DirectDraw; JPEG image compression algorithm

目录
1 绪论 1
1.1 课题来源 1
1.2 课题目的和意义 1
1.3 课题解决的问题 2
1.4 本文的组织结构 2
2 相关技术知识介绍 3
2.1镜像驱动截屏技术 3
2.1.1 Mirror driver简介 3
2.1.2 Mirror driver驱动原理 5
2.2 位图文件格式 6
2.2.1 位图文件头 7
2.2.2 位图信息头 7
2.2.3 调色板 8
2.2.4 图像数据 9
2.3 图像压缩算法 9
2.3.1 图像压缩步骤 9
2.3.1 普通JPEG压缩算法 11
3 系统模块功能的实现 13
3.1安装Mirror driver 13
3.2 屏幕捕获 16
3.2.1 全屏捕获 16
3.2.2 位图的保存 17
3.2.3 区域截屏 19
3.2.4 截屏效果 20
3.3 位图压缩及显示 22
3.3.1 按指定质量将位图压缩为JPEG图像 22
3.3.2 压缩后图像的显示 27
3.3.3 位图压缩效果 28
3.4 图像的连续显示 29
3.4.1 使用定时器SetTimer 29
3.4.2 使用Sleep( time )函数 30
3.5 图像连续显示效果 32
结论 34
致谢 35
参考文献 36

镜像驱动截屏技术在数字语音教室中的应用由免费论文网（www.jaoyuw.com）会员上传。