C# - 减小图像大小的最佳方法是什么（用于通过 tcp 进行流式传输）
我编写了一个将实时屏幕流式传输到远程应用程序的应用程序。它抓取屏幕（将图像大小调整为 640x480），然后使用 GIF 压缩（使用 System.Drawing）来压缩图像，将其保存到 byte[] 数组中并传输它到另一个应用程序。
问题是我获得的图像约为 50KB，这意味着在 30FPS 下每秒需要传输 1.5MB 的数据。目前我只能获得 8-10 FPS。我知道可以通过某种方式解决这个问题。也许使用 Flash 视频使用的技术？
I've written an application that streams live screen to a remote app. It grabs the screen (resizes the image to 640x480) and then compresses the image using GIF compression (using System.Drawing), saves it into a byte[] array and transfers it to the other app.
The problem is that the image I get is about 50KB which means that at 30FPS it would require 1.5MB of data transferred each second. At the moment I only get 8-10 FPS. I know it's possible to solve this somehow. Maybe using the technique that flash videos use?
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（4）
就我个人而言，我建议使用 VNCSharp - 它会为您完成大部分繁重的工作。有些人可能会说，重新编写这个代码太疯狂了。
如果没有，那么流式传输图像就会浪费带宽 - 您需要有效地构建视频流并传输它。
Personally I'd recommend using VNCSharp - it will do most of the heavy lifting for you. Some might say that it'd be madness to code this up again.
If not then streaming images is a waste of bandwidth - you need to effectively build a video stream and transmit that.
由于您不需要动画并且希望保持无损压缩，因此使用 PNG 而不是 GIF 可以获得更好的压缩效果（而且 PNG 是无专利的）。 据此可节省 10% 到 30% 之间。
Since you don't need animation and want to stay with loss-less compression you would get somewhat better compression with PNG instead of GIF (and PNG is patent-free). According to this the savings is between 10 to 30%.
我认为使用逐个屏幕捕获并不是获得实时屏幕流的好方法。视频格式通常假设在许多帧之间有一小部分区域会发生变化。另一方面，您需要做更多工作才能从屏幕上捕获视频。
您可以从以下文章开始：
http:// /betterlogic.com/roger/2010/07/list-of-available-directshow-screen-capture-filters/
http://www.codeproject.com/KB/dialog/screencap.aspx
I think use screen by screen capture isn´t a good approach to get a live screen streaming. Video formats, usually, assumes that between a lot of frames you have a small couple of areas that changes. In the other hand you´ll need to work a bit more to capture a video from the screen.
You can start from these articles:
http://betterlogic.com/roger/2010/07/list-of-available-directshow-screen-capture-filters/
http://www.codeproject.com/KB/dialog/screencap.aspx
您最好压缩视频流，而不是压缩图像。这就是视频编解码器实现高压缩的方式：通过利用流中连续图像的相似性。
如果你一张一张地压缩图像，你就会失去这种性能优势，并且会在带宽上产生巨大的差异。
Rather than compressing images, you'd better compress video streams. This is how video codec achieve high compression : by exploiting similarities into consecutives images in the stream.
If you compress your images one by one, you loose this performance advantage, and it makes a huge difference in bandwidth.