从B站上看到的一个不错的Golang项目，避免自己遗忘，记录一下
参考B站视频：https://www.bilibili.***/video/BV15L41187LY?p=13&vd_source=bdee36feb4edbfc7100eaadac4772a5e
部署部分（这部分自己探索的）
需要一台云服务器（我是阿里云），并安装docker（使用Dockerfile）
一、上传项目到阿里云服务器
依旧是xshell，注意把Go项目放进压缩包，不能直接上传项目文件夹。
具体上传步骤可以参考我之前的文章，部署springboot项目那篇。
二、部署
1.解压项目
我的Go项目名为ocr
unzip ocr.zip
2.使用Docker
mkdir ocr
cd ocr
vi Dockerfile
Dockerfile内容如下：
# 使用官方的 Go 镜像作为基础镜像
FROM golang:latest
# 设置工作目录
WORKDIR /app
# 复制 go.mod 和 go.sum 文件并下载依赖
COPY go.mod .
COPY go.sum .
RUN go env -w GOPROXY=https://proxy.golang.***.***,https://goproxy.***,direct
RUN go mod download
# 将本地的代码复制到容器中的工作目录
COPY . .
# 编译 Go 项目
RUN go build -o main .
# 暴露应用程序的端口
EXPOSE 80
EXPOSE 8080
# 运行应用程序
CMD ["./main"]
接着执行：
docker build -t ocr .
docker run -p 8080:8080 -p 80:80 ocr
执行完毕查看docker 的images和containers：
—
三：测试
使用Apifox进行测试
测试对普通印刷物品
使用imageUrl
访问：http://47.94.246.106:80/ocr/general/basic，post请求，body体传输必要参数
服务器会返回一系列json格式数据。
DetectedText : 所识别的文字
Confidence : 置信度
Polygon : 检测到的四个角的坐标
ItemPolygon : 文本行在旋转纠正之后的图像中的像素坐标，表示为（左上角 x, 左上角 y，宽 width，高 height）。
{
"Response": {
"TextDetections": [
{
"DetectedText": "TetraP",
"Confidence": 100,
"Polygon": [
{
"X": 183,
"Y": 43
},
{
"X": 206,
"Y": 10
},
{
"X": 214,
"Y": 16
},
{
"X": 191,
"Y": 49
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":1}}",
"ItemPolygon": {
"X": 477,
"Y": 173,
"Width": 40,
"Height": 10
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "生牛乳",
"Confidence": 100,
"Polygon": [
{
"X": 141,
"Y": 296
},
{
"X": 220,
"Y": 260
},
{
"X": 235,
"Y": 293
},
{
"X": 156,
"Y": 330
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":8}}",
"ItemPolygon": {
"X": 249,
"Y": 290,
"Width": 86,
"Height": 36
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "特仑苏",
"Confidence": 100,
"Polygon": [
{
"X": 227,
"Y": 133
},
{
"X": 307,
"Y": 34
},
{
"X": 342,
"Y": 63
},
{
"X": 262,
"Y": 161
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":2}}",
"ItemPolygon": {
"X": 431,
"Y": 262,
"Width": 127,
"Height": 45
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "全脂灭菌乳",
"Confidence": 100,
"Polygon": [
{
"X": 159,
"Y": 338
},
{
"X": 291,
"Y": 279
},
{
"X": 305,
"Y": 310
},
{
"X": 173,
"Y": 369
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":7}}",
"ItemPolygon": {
"X": 226,
"Y": 329,
"Width": 144,
"Height": 34
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "贮存条件:常温密闭保存。",
"Confidence": 100,
"Polygon": [
{
"X": 258,
"Y": 238
},
{
"X": 377,
"Y": 86
},
{
"X": 393,
"Y": 99
},
{
"X": 274,
"Y": 251
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":3}}",
"ItemPolygon": {
"X": 365,
"Y": 349,
"Width": 193,
"Height": 20
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "n",
"Confidence": 50,
"Polygon": [
{
"X": 156,
"Y": 390
},
{
"X": 171,
"Y": 418
},
{
"X": 154,
"Y": 428
},
{
"X": 138,
"Y": 399
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":18}}",
"ItemPolygon": {
"X": 182,
"Y": 358,
"Width": 31,
"Height": 19
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "GB 25190",
"Confidence": 100,
"Polygon": [
{
"X": 178,
"Y": 377
},
{
"X": 300,
"Y": 324
},
{
"X": 314,
"Y": 356
},
{
"X": 192,
"Y": 410
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":18}}",
"ItemPolygon": {
"X": 206,
"Y": 368,
"Width": 133,
"Height": 34
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "开启前，无需冷藏",
"Confidence": 100,
"Polygon": [
{
"X": 274,
"Y": 250
},
{
"X": 359,
"Y": 142
},
{
"X": 375,
"Y": 155
},
{
"X": 290,
"Y": 263
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":4}}",
"ItemPolygon": {
"X": 365,
"Y": 369,
"Width": 137,
"Height": 20
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "开启后，请立即饮用。",
"Confidence": 100,
"Polygon": [
{
"X": 290,
"Y": 262
},
{
"X": 386,
"Y": 138
},
{
"X": 403,
"Y": 150
},
{
"X": 307,
"Y": 275
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":9}}",
"ItemPolygon": {
"X": 365,
"Y": 389,
"Width": 156,
"Height": 20
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "6个月",
"Confidence": 100,
"Polygon": [
{
"X": 196,
"Y": 421
},
{
"X": 263,
"Y": 390
},
{
"X": 278,
"Y": 422
},
{
"X": 210,
"Y": 453
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":24}}",
"ItemPolygon": {
"X": 181,
"Y": 409,
"Width": 73,
"Height": 35
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "请勿连同包装在微波炉中加热。",
"Confidence": 100,
"Polygon": [
{
"X": 307,
"Y": 273
},
{
"X": 438,
"Y": 97
},
{
"X": 456,
"Y": 110
},
{
"X": 325,
"Y": 286
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":7}}",
"ItemPolygon": {
"X": 366,
"Y": 409,
"Width": 219,
"Height": 22
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "贝包装喷码",
"Confidence": 100,
"Polygon": [
{
"X": 218,
"Y": 458
},
{
"X": 338,
"Y": 404
},
{
"X": 352,
"Y": 436
},
{
"X": 233,
"Y": 490
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":26}}",
"ItemPolygon": {
"X": 164,
"Y": 448,
"Width": 131,
"Height": 34
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "料",
"Confidence": 100,
"Polygon": [
{
"X": 376,
"Y": 245
},
{
"X": 388,
"Y": 227
},
{
"X": 403,
"Y": 237
},
{
"X": 391,
"Y": 255
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":10}}",
"ItemPolygon": {
"X": 430,
"Y": 448,
"Width": 21,
"Height": 18
},
"Words": [],
"WordCoordPoint": []
},
{
"DetectedText": "生牛乳",
"Confidence": 100,
"Polygon": [
{
"X": 388,
"Y": 225
},
{
"X": 420,
"Y": 183
},
{
"X": 436,
"Y": 195
},
{
"X": 403,
"Y": 238
}
],
"AdvancedInfo": "{\"Parag\":{\"ParagNo\":5}}",
"ItemPolygon": {
"X": 453,
"Y": 446,
"Width": 52,
"Height": 20
},
"Words": [],
"WordCoordPoint": []
},
... ...
... ...
... ...
... ...
测试对普通印刷物品，使用image_file
访问：http://47.94.246.106:80/ocr/general/basic，post请求，body体传输必要参数
识别成功。
测试健康码
使用imageUrl
访问：http://47.94.246.106:80/ocr/general/healthcode，post请求，body体传输必要参数
服务器响应返回json数据：
{
"Response": {
"Name": "郭**",
"IDNumber": "",
"Time": "09-29 11:17:41",
"Color": "绿色",
"TestingInterval": "",
"TestingResult": "阴性",
"TestingTime": "2021-08-27 17:00",
"Va***ination": "暂未查询到数据",
"SpotName": "",
"Va***inationTime": "",
"RequestId": "45fe12a8-9018-46ae-874b-07ba53703475"
}
}
使用image_file
访问：http://47.94.246.106:80/ocr/general/healthcode，post请求，body体传输必要参数
识别成功
四、还有几件事
1、整个识别逻辑在腾讯云那里，还有很多可以识别的东西，比如车牌识别等等， 识别质量还有高精度识别等。
2、写完一个识别，基于固定套路很容易扩展，基本当时写完文字识别，自己调了差不多5分钟，就把健康码识别搞定了。
3、现在只是返回了Json数据，如何实现友好的前端交互渲染，额目前先不考虑，相比之下自己更喜欢后端，而且想把Go学好。
4、记录一下，免得哪天自己忘了，过来能回忆回忆。
5、还有就是，腾讯云的这个识别，刚免费开通时提供1000次免费次数，之后需要购买。
3月12日上午，使用Gin框架完成了项目结构的重写。
