车牌识别 [英] vehicle number plate recognition

问题描述

我正在做一个人工智能项目.
该项目的定义是从车辆(无论是汽车还是踏板车)中提取车牌.
首先是将彩色图像转换为二值化

任何人都可以帮忙吗?

I am doing a project in artificial intelligence.
The project definition is to extract the number plate from a vehicle (either a car or scooter anything).
The fisrt step is to covert the color image into binarisation

Can anyone Help?

推荐答案

这不是真的AI，除非您要进入模糊逻辑来确定字符是什么.无论如何，这些步骤是:

0.图像捕获.
1.识别车牌-查看形状识别算法.
2.提取注册号.此处的关键词是光学字符识别[OCR]，并且已有用于此的库.问Google博士.

如果您可以编写自己的OCR例程-不管它是否适用于车牌号-您都会在线获得最高评分，因为它刻苦钻研.祝你好运.

This isn''t really AI unless you are going to get into fuzzy logic to decide what the characters are. Anyhow, the steps are:

0. Image capture.
1. Identify the number plate - look at shape recognition algorithms for this.
2. Extracting the registration number. The key word here is Optical Character Recognition [OCR] and there are existing libraries for this; ask Dr Google.

If you can write your own OCR routine - irrespective of if it''s for number plates or not - you''ll be on line for top marks as it''s frogging hard. Good luck.

这篇关于车牌识别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！