使用C#版Tesseract库

上一篇介绍了Tesseract库的使用(OCR库Tesseract初探),文末提到了Tesseract是用c/c++开发的,也有C#的开源版本,本篇介绍一下如何使用C#版的Tesseract。

C#版本源码下载地址:https://github.com/charlesw/tesseract

其实在vs中可以直接用NuGet工具进行下载:

打开nuget,搜索tesseract,点安装即可。

 

源码是vs2015编译的,需要安装vs2015以上版本。

打开项目后如:

我们再添加一个winform项目,画界面如:

实现点击“选择需要识别的图片”,打开一张图片,调用算法并显示结果。比较简单。源码如下:

using System;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Windows.Forms;
using Tesseract;

namespace TesseractDemo
{
    public partial class Form1 : Form
    {
        public Form1()
        {
            InitializeComponent();
        }
        //选图片并调用ocr识别方法
        private void btnRec_Click(object sender, EventArgs e)
        {
            //openFileDialog1.Filter = "";
            if (openFileDialog1.ShowDialog() == DialogResult.OK)
            {
                var imgPath = openFileDialog1.FileName;
                pictureBox1.Image=Image.FromFile(imgPath);
                string strResult = ImageToText(imgPath);
                if (string.IsNullOrEmpty(strResult))
                {
                    txtResult.Text = "无法识别";
                }
                else
                {
                    txtResult.Text = strResult;
                }
            }
        }
        //调用tesseract实现OCR识别
        public string ImageToText(string imgPath)
        {
            using (var engine = new TesseractEngine("tessdata", "eng", EngineMode.Default))
            {
                using (var img = Pix.LoadFromFile(imgPath))
                {
                    using (var page = engine.Process(img))
                    {
                        return page.GetText();
                    }
                }
            }
        }
    }
}

 有一点要注意的是,tesseract的识别语言包要自己下载后包含到项目里面,并设置为始终复制,或者直接把这个文件包放到运行程序目录(bin\debug)下:

eng是英文字符的意思,要识别其他语言字符,需要自己下载:

Tesseract has unicode (UTF-8) support, and can recognize more than 100 languages "out of the box".

这个库支持100种语言的识别

字库下载地址为:https://github.com/tesseract-ocr/tessdata

用OpencvSharp先降噪再调OCR识别:

//用opencv进行降噪处理再ocr识别
        private void button3_Click(object sender, EventArgs e)
        {
            //从网上读取一张图片
            string imgUrl = "https://service.cheshi.com/user/validate/validatev3.php";
            MemoryStream ms = ReadImgFromWeb(imgUrl);
            Image img = Image.FromStream(ms);
            pictureBox1.Image = img;

            //降噪
            Mat simg = Mat.FromStream(ms, ImreadModes.Grayscale);
            Cv2.ImShow("Input Image", simg);
            //阈值操作 阈值参数可以用一些可视化工具来调试得到
            Mat ThresholdImg = simg.Threshold(29, 255, ThresholdTypes.Binary);
            Cv2.ImShow("Threshold", ThresholdImg);
            Cv2.ImWrite("d:\\img.png", ThresholdImg);
            
            textBox1.Text= ImageToText("d:\\img.png");
        }
        
        /// <summary>
        /// 从网上读取一张图片
        /// </summary>
        /// <param name="Url"></param>
        public MemoryStream ReadImgFromWeb(string Url)
        {
            HttpWebRequest request = (HttpWebRequest)WebRequest.Create(Url);
            request.Credentials = CredentialCache.DefaultCredentials; // 添加授权证书
            request.UserAgent = "Microsoft Internet Explorer";
            WebResponse response = request.GetResponse();
            Stream s = response.GetResponseStream();
            byte[] data = new byte[1024];
            int length = 0;
            MemoryStream ms = new MemoryStream();
            while ((length = s.Read(data, 0, data.Length)) > 0)
            {
                ms.Write(data, 0, length);
            }
            ms.Seek(0, SeekOrigin.Begin);
            //pictureBox1.Image = Image.FromStream(ms);
            return ms;
        }

请自行用NuGet程序下载opencvsharp3.0库,参考https://www.cnblogs.com/tuyile006/p/10819570.html

posted @ 2019-04-26 09:54 小y 阅读(...) 评论(...) 编辑 收藏