是应该在客户端还是在服务器端处理OCR？

I am trying build an Android app which should be able to take a picture of a figure. The figure in this picture should be extracted with OCR and stored on a web service. Would it be better to do the OCR-processing on the client or on the server side?

我正在尝试构建一个能够拍摄人物照片的Android应用程序。应使用OCR提取此图中的数字并将其存储在Web服务中。在客户端或服务器端进行OCR处理会更好吗?

In my opinion the advantages for the client OCR-processing outweigh the advantages of the server side processing.

在我看来,客户端OCR处理的优势超过了服务器端处理的优势。

Advantages of client-side OCR:

客户端OCR的优点:

Minimum data traffic

最小数据流量

Works offline
Faster

Can you give me any suggestions on whether the client or server-side processing is better or any points I am missing?

您能否就客户端或服务器端处理是否更好或我缺少的任何一点给出任何建议?

1 个解决方案

#1

Have a look at this:

看看这个:

https://github.com/rmtheis/android-ocr

Some pros and cons when using local and remote image elaboration. I suggest to give a try to this project and find out performance. Obviously, web-server is fast and powerfull but forces users to be always on line when using your app, but as I understand your app works on-line, so is it a real problem?

使用本地和远程图像细化时的一些优点和缺点。我建议试试这个项目并找出性能。显然,网络服务器速度快,功能强大,但迫使用户在使用你的应用程序时总是在线,但据我所知,你的应用程序在线工作,这是一个真正的问题吗?

A lot of Google apps, use server power to make computation of picture and voice recordings.

许多Google应用程序使用服务器电源来计算图片和录音。

Moving the computation to server also allow you to make improvements to the OCR, so no need to update the app.

将计算移动到服务器还允许您对OCR进行改进,因此无需更新应用程序。

Good luck!

#1