连接的组件标签,用于C#中的文本检测 [英] Connected component labeling for text detection in C#

查看:100
本文介绍了连接的组件标签,用于C#中的文本检测的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我正在做一个项目,我必须在自然图像中检测文本,在最初的边缘图生成阶段之后,必须对连接的组件进行标注以识别文本组件.在C#中是否有任何为文本检测完成连接组件标签的示例?

请不要粗鲁的回答,这是我第一次必须实现这样的事情.

通过自然图像,我的意思是来自现实世界的图像,例如招牌.

I''m doing project where I have to detect text in natural images, after the initial edge map generation stage, the connected component labeling must be done to identify text components. Are there any examples of connected component labeling done for text detection in C#?

Please no rude answers, this is the first time I''ve had to implement something like this.

By natural images I meant images from real world like signboards for instance.

推荐答案

我认为自然的意思是现实世界".这很艰难,而且我不知道有人在不使用预建组件的情况下进行此操作.
I assume natural means ''real world''. This is tough and I don''t know of anyone doing this without using a prebuilt component to do it. Which usually means a commercial one.


有一个名为Leadtools的工具包可以帮助您满足需求.您可以使用它对图像进行一些处理(即对图像进行二值化处理,从图像中去除不需要的点,将其转换为1位黑白等),然后将其传递给其OCR引擎.
您可以检查他们的网站以获取更多信息,或与他们的支持团队联系.
There is a toolkit named leadtools that could help you doing your requirements. You can use it to do some processing on the image (i.e. binrize it, remove the unwanted dots from the image, convert it to 1-bit black & white, etc), and then pass it to their OCR engine.
You can check their website for more information, or contact their support team.


这篇关于连接的组件标签,用于C#中的文本检测的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆