NodeJS 搭建图形识别功能

柔光的暖阳◎ 2023-10-18 16:24 113阅读 0赞

第一步：创建项目存储路径(C:\\node\_workspace),创建NodeJs 项目(image\_reco) 并初始化

cd C:\\node\_workspace

mkdir image\_reco

cnpm init

第二步：图形识别依赖模块介绍和本地安装

[node-images][]: Node.js 轻量级跨平台图像编解码库  
[tesseract][].js: 纯 JS 的 OCR 库支持 62 种语言。

本地安装:[node-images][] 和 [tesseract][].js模块

C:\node_workspace\image_reco>npm install tesseract.js
    npm notice created a lockfile as package-lock.json. You should commit this file.
    npm WARN image_reco@1.0.0 No description
    npm WARN image_reco@1.0.0 No repository field.
    
    + tesseract.js@1.0.14
    added 23 packages in 9.246s
    
    C:\node_workspace\image_reco>npm install images
    
    > images@3.0.2 install C:\node_workspace\image_reco\node_modules\images
    > node ./scripts/install.js
    
    Downloading binary from https://github.com/zhangyuanwei/node-images/releases/download/v3.0.2/win32-x64-57_binding.node
    Download complete..] - :
    Binary saved to C:\node_workspace\image_reco\node_modules\images\vendor\win32-x64-57\binding.node
    Caching binary to C:\Users\zzg\AppData\Roaming\npm-cache\images\3.0.2\win32-x64-57_binding.node
    npm WARN image_reco@1.0.0 No description
    npm WARN image_reco@1.0.0 No repository field.
    
    + images@3.0.2
    added 97 packages in 15.86s

编辑package.json 文件，添加第三方模块

{
      "name": "image_reco",
      "version": "1.0.0",
      "description": "",
      "main": "app.js",
      "scripts": {
        "test": "echo \"Error: no test specified\" && exit 1"
      },
      "author": "",
      "license": "ISC",
      "dependencies": {
        "images": "^3.0.2",
        "tesseract.js": "^1.0.14"
      }
    }

第三步：项目根路径编写app.js 文件，编辑内容如下：

"use strict"
    var images = require('images')
    var Tesseract = require('tesseract.js');
    var request = require('request');
    var fs = require('fs');
    // 将图片下载到本地
    function downloadFile(uri, filename, callback) {
        var stream = fs.createWriteStream(filename);
        request(uri).pipe(stream)
            .on('close', function () {
                callback();
            });
    }
    // 识别图片
    function recognize(filePath, callback) {
        // 图片放大
        images(filePath)
            .size(90)
            .save(filePath);
        // 识别
        Tesseract
            .recognize(filePath, {
                lang: 'eng', // 语言选英文
                tessedit_char_blacklist: 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
                    //因为是数字验证码，排除字母
            })
            .then((result) => {
                callback(result.text);
            });
    }
    function getVcode() {
        var url = 'https://ww1.sinaimg.cn/large/8c9b876fly1fe0bvsibibj201a00p07l.jpg';
        var filename = 'vcode.png';
        // 先下载下来，再识别
        downloadFile(url, filename, function () {
            recognize(filename, function (txt) {
                console.log('识别结果: ' + txt);
            });
        });
    }
    getVcode();

第四步：安装项目依赖，执行图片识别app.js 代码

C:\node_workspace\image_reco>cnpm install
    √ Installed 2 packages
    √ Linked 0 latest versions
    √ Run 0 scripts
    √ All packages installed (used 34ms(network 22ms), speed 0B/s, json 0(0B), tarball 0B)
    
    C:\node_workspace\image_reco>node app.js
    pre-main prep time: 37 ms
    识别结果: 3128

项目结构：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3pob3V6aGl3ZW5nYW5n_size_16_color_FFFFFF_t_70][]

[node-images]: https://github.com/zhangyuanwei/node-images
[tesseract]: https://github.com/naptha/tesseract.js
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3pob3V6aGl3ZW5nYW5n_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20190329175713531.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3pob3V6aGl3ZW5nYW5n,size_16,color_FFFFFF,t_70