CSharp SDK_人脸识别服务 FRS_SDK参考_SDK（旧版）_华为云

(1)

SDK 参考

文档版本 01

发布日期 2022-02-15

(2)

非经本公司书面许可，任何单位和个人不得擅自摘抄、复制本文档内容的部分或全部，并不得以任何形式传播。

商标声明

和其他华为商标均为华为技术有限公司的商标。

本文档提及的其他所有商标或注册商标，由各自的所有人拥有。

注意

您购买的产品、服务或特性等应受华为公司商业合同和条款的约束，本文档中描述的全部或部分产品、服务或特性可能不在您的购买或使用范围之内。除非合同另有约定，华为公司对本文档内容不做任何明示或暗示的声明或保证。

由于产品版本升级或其他原因，本文档内容会不定期进行更新。除非另有约定，本文档仅作为使用指导，本文档中的所有陈述、信息和建议不构成任何明示或暗示的担保。

华为技术有限公司

地址：深圳市龙岗区坂田华为总部办公楼邮编：518129 网址：

https://www.huawei.com

客户服务邮箱：

[email protected]

客户服务电话：4008302118

(3)

1 ^简介

人脸识别服务软件开发工具包（FRS SDK）是对人脸识别服务提供的REST API进行的封装，以简化用户的开发工作。FRS SDK目前支持Java、Python、Go、.NET、

Node.js、PHP、C++版本。

接口与 API 对应关系

人脸识别接口与API对应关系请参见表1-1。

表1-1 接口与 API 对应关系表

接口 API

人脸检测 POST /v2/{project_id}/face-detect 人脸比对 POST /v2/{project_id}/face-compare 人脸搜索 POST /v2/{project_id}/face-sets/

{face_set_name}/search

活体检测动作活体检测 POST /v1/{project_id}/live-detect 静默活体检测 POST /v1/{project_id}/live-detect-face 人脸库资源管

理

创建人脸库 POST /v2/{project_id}/face-sets 查询所有人脸

库

GET /v2/{project_id}/face-sets

查询人脸库 GET /v2/{project_id}/face-sets/{face_set_name}

删除人脸库 DELETE /v2/{project_id}/face-sets/

{face_set_name}

人脸资源管理添加人脸 POST /v2/{project_id}/face-sets/

{face_set_name}/faces

查询人脸 GET /v2/{project_id}/face-sets/

{face_set_name}/faces?offset=xxx&limit=xxx

(5)

接口 API

更新人脸 PUT /v2/{project_id}/face-sets/

{face_set_name}/faces

删除人脸 DELETE /v2/{project_id}/face-sets/

{face_set_name}/faces?field_name=field_value 批量删除人脸 DELETE /v2/{project_id}/face-sets/

{face_set_name}/faces/batch

(6)

2 ^{Java SDK}

本章节介绍人脸识别服务Java SDK，您可以参考本章节进行快速集成开发。

准备工作

● 已注册华为云帐号，并完成实名认证，帐号不能处于欠费或冻结状态。

● 已开通人脸识别服务。如未开通，请登录人脸识别管理控制台开通所需服务。

● 已具备开发环境，支持Java JDK 1.8 及其以上版本。

● 登录“我的凭证 > 访问秘钥”页面，获取Access Key（AK）和Secret Access Key

（SK）。

● 登录“我的凭证”页面，获取“IAM用户名”“、帐号名”以及待使用区域的

“项目ID”。调用服务时会用到这些信息，请提前保存。

本样例以“华北-北京四”区域为例，获取对应的项目ID（project_id）。

图2-1 我的凭证

安装 SDK

推荐您通过Maven方式获取和安装SDK，首先需要在您的操作系统中下载并安装

Maven，安装完成后您只需要在Java项目的pom.xml文件中加入相应的依赖项即可。

使用SDK前，需要安装“huaweicloud-sdk-core”和“huaweicloud-sdk-frs”。SDK 的版本信息请参见SDK中心。

<groupId>com.huaweicloud.sdk</groupId>

<artifactId>huaweicloud-sdk-core</artifactId>

</dependency>

(7)

<groupId>com.huaweicloud.sdk</groupId>

<artifactId>huaweicloud-sdk-frs</artifactId>

</dependency>

开始使用

详细的SDK介绍，使用异步客户端，配置日志请参见SDK中心、Java SDK使用指导、

Java SDK使用视频。

1. 导入依赖模块

import com.huaweicloud.sdk.core.auth.ICredential;

import com.huaweicloud.sdk.core.auth.BasicCredentials;

import com.huaweicloud.sdk.core.exception.ConnectionException;

import com.huaweicloud.sdk.core.exception.RequestTimeoutException;

import com.huaweicloud.sdk.core.exception.ServiceResponseException;

//导入v2版本sdk

import com.huaweicloud.sdk.frs.v2.region.FrsRegion;

import com.huaweicloud.sdk.frs.v2.*;

import com.huaweicloud.sdk.frs.v2.model.*;

2. 配置认证信息

配置AK、SK信息。华为云通过AK识别用户的身份，通过SK对请求数据进行签名验证，用于确保请求的机密性、完整性和请求者身份的正确性。AK、SK获取方法请参见准备工作。

public static ICredential getCredential(String ak, String sk) { return new BasicCredentials()

.withAk(ak) .withSk(sk);

}

3. 初始化客户端指定region方式

public static FrsClient getClient(Region region, ICredential auth) { // 初始化人脸识别服务的客户端

return FrsClient.newBuilder() .withCredential(auth)

.withRegion(region) // 选择服务所在区域 FrsRegion.valueOf("cn-north-4") .build();

}

服务部署区域请参见终端节点。

4. 发送请求并查看响应

// 以调用人脸检测 DetectFaceByBase64 接口为例

DetectFaceByBase64Request request = new DetectFaceByBase64Request();

FaceDetectBase64Req body = new FaceDetectBase64Req();

body.withImageBase64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

request.withBody(body);

DetectFaceByBase64Response response = client.detectFaceByBase64(request);

System.out.println(response.toString());

说明

使用人脸比对SDK时，image1、image2参数需为相同类型，即同为url、base64或file。

5. 异常处理

表2-1 异常处理

一级分类一级分类说明二级分类二级分类说明

ConnectionExcepti

on 连接类异常 HostUnreachableE

xception 网络不可达、被拒绝。

(8)

一级分类一级分类说明二级分类二级分类说明 SslHandShakeExce

ption SSL认证异常。

RequestTimeoutEx

ception 响应超时异常 CallTimeoutExcept

ion 单次请求，服务器

处理超时未返回。

RetryOutageExcep

tion 在重试策略消耗完

成后，仍无有效的响应。

ServiceResponseEx

ception 服务器响应异

常 ServerResponseEx

ception 服务端内部错误，

Http响应码：

[500,]。

ClientRequestExce

ption 请求参数不合法，

Http响应码：[400, 500)

DetectFaceByBase64Request request = new DetectFaceByBase64Request();

try {

DetectFaceByBase64Response response = client.detectFaceByBase64(request);

} catch (ConnectionException e) { e.printStackTrace();

} catch (RequestTimeoutException e) { e.printStackTrace();

} catch (ServiceResponseException e) { e.printStackTrace();

System.out.println(e.getHttpStatusCode());

System.out.println(e.getErrorCode());

System.out.println(e.getErrorMsg());

}

SDK demo 代码解析

● 人脸检测

DetectFaceByBase64Request detectRequest = new DetectFaceByBase64Request();

FaceDetectBase64Req faceDetectBase64Req = new FaceDetectBase64Req();

faceDetectBase64Req.withImageBase64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

faceDetectBase64Req.withAttributes("2");

detectRequest.setBody(faceDetectBase64Req);

try {

DetectFaceByBase64Response detectResponse = client.detectFaceByBase64(detectRequest);

System.out.println(detectResponse.toString());

}

● 人脸比对

CompareFaceByBase64Request compareRequest = new CompareFaceByBase64Request();

FaceCompareBase64Req faceCompareBase64Req = new FaceCompareBase64Req();

faceCompareBase64Req.withImage1Base64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

faceCompareBase64Req.withImage2Base64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

(9)

compareRequest.withBody(faceCompareBase64Req);

try {

CompareFaceByBase64Response compareResponse = client.compareFaceByBase64(compareRequest);

System.out.println(compareResponse.toString());

}

● 人脸搜索

SearchFaceByBase64Request searchRequest = new SearchFaceByBase64Request();

searchRequest.withFaceSetName("face_set_name");

FaceSearchBase64Req faceSearchBase64Req = new FaceSearchBase64Req();

List<Map<String, String>> listbodySort = new ArrayList<>();

Map<String, String> map = new HashMap<>();

map.put("timestamp","asc");

listbodySort.add(map);

List<String> listbodyReturnFields = new ArrayList<>();

listbodyReturnFields.add("timestamp");

listbodyReturnFields.add("id");

faceSearchBase64Req.withSort(listbodySort);

faceSearchBase64Req.withReturnFields(listbodyReturnFields);

faceSearchBase64Req.withImageBase64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

searchRequest.withBody(faceSearchBase64Req);

try {

SearchFaceByBase64Response searchResponse = client.searchFaceByBase64(searchRequest);

System.out.println(searchResponse.toString());

}

● 创建人脸库

CreateFaceSetRequest createFaceSetRequest = new CreateFaceSetRequest();

CreateFaceSetReq createFaceSetReq = new CreateFaceSetReq();

createFaceSetReq.withFaceSetName("face_set_name");

Map<String, TypeInfo> stringTypeInfoMap = new HashMap<>();

TypeInfo typeInfo = new TypeInfo();

typeInfo.withType("long");

stringTypeInfoMap.put("timestamp", typeInfo);

createFaceSetReq.withExternalFields(stringTypeInfoMap);

createFaceSetRequest.withBody(createFaceSetReq);

try {

CreateFaceSetResponse createFaceSetResponse = client.createFaceSet(createFaceSetRequest);

System.out.println(createFaceSetResponse.toString());

}

● 查询人脸库

(10)

ShowFaceSetRequest showFaceSetRequest = new ShowFaceSetRequest();

showFaceSetRequest.withFaceSetName("face_set_name");

try {

ShowFaceSetResponse showFaceSetResponse = client.showFaceSet(showFaceSetRequest);

System.out.println(showFaceSetResponse.toString());

}

● 查询所有人脸库

ShowAllFaceSetsRequest showAllFaceSetsRequest = new ShowAllFaceSetsRequest();

try {

ShowAllFaceSetsResponse showAllFaceSetsResponse = client.showAllFaceSets(showAllFaceSetsRequest);

System.out.println(showAllFaceSetsResponse.toString());

}

● 删除人脸库

DeleteFaceSetRequest deleteFaceSetRequest = new DeleteFaceSetRequest();

deleteFaceSetRequest.withFaceSetName("face_set_name");

try {

DeleteFaceSetResponse deleteFaceSetResponse = client.deleteFaceSet(deleteFaceSetRequest);

System.out.println(deleteFaceSetResponse.toString());

}

● 添加人脸

//add face

AddFacesByBase64Request addFacesByBase64Request = new AddFacesByBase64Request();

addFacesByBase64Request.withFaceSetName("face_set_name");

AddFacesBase64Req addFacesBase64Req = new AddFacesBase64Req();

addFacesBase64Req.withExternalFields("{\"timestamp\":12}");

addFacesBase64Req.withImageBase64("9j/4AAQSkZJRgABAQAAAQABAAD...");

addFacesByBase64Request.withBody(addFacesBase64Req);

try {

AddFacesByBase64Response addFacesByBase64Response = client.addFacesByBase64(addFacesByBase64Request);

System.out.println(addFacesByBase64Response.toString());

(11)

}

● 删除人脸

//delete face by faceId

DeleteFaceByFaceIdRequest deleteFaceByFaceIdRequest = new DeleteFaceByFaceIdRequest();

deleteFaceByFaceIdRequest.withFaceSetName("face_set_name");

deleteFaceByFaceIdRequest.withFaceId("iexEBb6t");

try {

DeleteFaceByFaceIdResponse deleteFaceByFaceIdResponse = client.deleteFaceByFaceId(deleteFaceByFaceIdRequest);

System.out.println(deleteFaceByFaceIdResponse.toString());

}

//delete face by externalImageId

DeleteFaceByExternalImageIdRequest deleteFaceRequest = new DeleteFaceByExternalImageIdRequest();

deleteFaceRequest.withFaceSetName("face_set_name");

deleteFaceRequest.withExternalImageId("iexEBb6t");

try {

DeleteFaceByExternalImageIdResponse response = client.deleteFaceByExternalImageId(deleteFaceRequest);

}

● 批量删除人脸

BatchDeleteFacesRequest batchDeleteFacesRequest = new BatchDeleteFacesRequest();

batchDeleteFacesRequest.withFaceSetName("face_set_name");

DeleteFacesBatchReq deleteFacesBatchReq = new DeleteFacesBatchReq();

deleteFacesBatchReq.withFilter("age:[20 TO 30]");

batchDeleteFacesRequest.withBody(deleteFacesBatchReq);

try {

BatchDeleteFacesResponse batchDeleteFacesResponse = client.batchDeleteFaces(batchDeleteFacesRequest);

System.out.println(batchDeleteFacesResponse.toString());

}

● 更新人脸

UpdateFaceRequest updateFaceRequest = new UpdateFaceRequest();

updateFaceRequest.withFaceSetName("face_set_name");

UpdateFaceReq updateFaceReq = new UpdateFaceReq();

updateFaceReq.withFaceId("iexEBb6t");

updateFaceRequest.withBody(updateFaceReq);

(12)

try {

UpdateFaceResponse response = client.updateFace(updateFaceRequest);

}

● 查询人脸

//show faces by faceId

ShowFacesByFaceIdRequest showFacesByFaceIdRequest = new ShowFacesByFaceIdRequest();

showFacesByFaceIdRequest.withFaceSetName("face_set_name");

showFacesByFaceIdRequest.withFaceId("iexEBb6t");

try {

ShowFacesByFaceIdResponse response = client.showFacesByFaceId(showFacesByFaceIdRequest);

}

//show faces by limit

ShowFacesByLimitRequest showFacesByLimitRequest = new ShowFacesByLimitRequest();

showFacesByLimitRequest.withFaceSetName("face_set_name");

showFacesByLimitRequest.withOffset(0);

showFacesByLimitRequest.withLimit(10);

try {

ShowFacesByLimitResponse showFacesByLimitResponse = client.showFacesByLimit(showFacesByLimitRequest);

System.out.println(showFacesByLimitResponse.toString());

}

● 动作活体检测

//detect live by base64

DetectLiveByBase64Request request = new DetectLiveByBase64Request();

LiveDetectBase64Req body = new LiveDetectBase64Req();

body.withActions("1,2,3,4");

body.withVideoBase64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

try {

DetectLiveByBase64Response response = client.detectLiveByBase64(request);

(13)

}

● 静默活体检测

//detect live face by base64

DetectLiveFaceByBase64Request request = new DetectLiveFaceByBase64Request();

LiveDetectFaceBase64Req body = new LiveDetectFaceBase64Req();

body.withImageBase64("/9j/4AAQSkZJRgABAQAAAQABAAD...");

try {

DetectLiveFaceByBase64Response response = client.detectLiveFaceByBase64(request);

}

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

看、在线咨询。

您只需要在API Explorer中修改接口参数，即可自动生成对应的代码示例。同时，可在集成开发环境CloudIDE中完成代码的构建、调试、运行等操作。

图2-2 API Explorer

(14)

3 Python SDK

本章节介绍人脸识别服务Python SDK，您可以参考本章节进行快速集成开发。

准备工作

● 已开通人脸识别服务。如未开通，请登录人脸识别管理控制台人脸识别管理控制台开通所需服务。

● 已具备开发环境，支持Python3及以上版本。

（SK）。

图3-1 我的凭证

安装 SDK

支持Python3及以上版本，执行python --version检查当前Python的版本信息。

使用SDK前，需要安装“huaweicloudsdkcore”和“huaweicloudsdkfrs”。

# 安装核心库

pip install huaweicloudsdkcore

# 安装FRS服务库

pip install huaweicloudsdkfrs

(15)

开始使用

详细的SDK介绍，使用异步客户端，配置日志请参见SDK中心、Python SDK使用指

导、Python SDK使用视频。

from huaweicloudsdkcore.auth.credentials import BasicCredentials from huaweicloudsdkcore.exceptions import exceptions

# 导入v2版本sdk

from huaweicloudsdkfrs.v2.region.frs_region import FrsRegion from huaweicloudsdkfrs.v2 import *

def GetCredential():

return BasicCredentials(ak, sk)

# 初始化人脸识别服务的客户端，并选择服务部署区域 def GetClient():

return FrsClient.new_builder(FrsClient) \ .with_credentials(credentials) \

.with_region(FrsRegion.value_of("cn-north-4")) \ .build()

# 以调用人脸检测 DetectFaceByBase64 接口为例 request = DetectFaceByBase64Request() request.body = FaceDetectBase64Req(

image_base64="/9j/4AAQSkZJRgABAQAAAQABAAD..."

)response = client.detect_face_by_base64(request) print(response)

说明

5. 异常处理

表3-1 异常处理

ConnectionExcepti

SslHandShakeExce

RequestTimeoutEx

RetryOutageExcep

(16)

一级分类一级分类说明二级分类二级分类说明 ServiceResponseEx

Http响应码：

[500,]。

ClientRequestExce

Http响应码：[400, 500)

try: request = DetectFaceByBase64Request() request.body = FaceDetectBase64Req(

)

response = client.detect_face_by_base64(request) print(response)

except exceptions.ClientRequestException as e:

print(e.status_code) print(e.request_id) print(e.error_code) print(e.error_msg)

SDK 代码解析

● 人脸检测

def detectFaceByBase64():

try:

request = DetectFaceByBase64Request() request.body = FaceDetectBase64Req(

image_base64="/9j/4AAQSkZJRgABAQAAAQABAAD...", attributes="2,4"

)

response = client.detect_face_by_base64(request) print(response)

● 人脸比对

def compareFaceByBase64():

try:

request = CompareFaceByBase64Request() request.body = FaceCompareBase64Req(

image1_base64="/9j/4AAQSkZJRgABAQAAAQABAAD...", image2_base64="/9j/4AAQSkZJRgABAQAAAQABAAD..."

)

response = client.compare_face_by_base64(request) print(response)

● 人脸搜索

def searchFaceByBase64():

try:

request = SearchFaceByBase64Request() request.face_set_name = "face_set_name"

listFaceSearchBase64ReqReturnFieldsbody = [ "timestamp"

(17)

]

request.body = FaceSearchBase64Req(

return_fields=listFaceSearchBase64ReqReturnFieldsbody, image_base64="/9j/4AAQSkZJRgABAQAAAQABAAD..."

)

response = client.search_face_by_base64(request) print(response)

● 创建人脸库

def createFaceSet():

try:

request = CreateFaceSetRequest() request.body = CreateFaceSetReq(

face_set_name="face_set_name",

external_fields={"timestamp": {"type": "long"}}

)

response = client.create_face_set(request) print(response)

● 查询人脸库

def showFaceSet():

try:

request = ShowFaceSetRequest() request.face_set_name = "face_set_name"

response = client.show_face_set(request) print(response)

def showAllFaceSet():

try:

request = ShowAllFaceSetsRequest() response = client.show_all_face_sets(request) print(response)

● 删除人脸库

def deleteFaceSet():

try:

request = DeleteFaceSetRequest() request.face_set_name = "face_set_name"

response = client.delete_face_set(request) print(response)

● 添加人脸

def addFacesByBase64():

try:

request = AddFacesByBase64Request()

(18)

request.face_set_name = "face_set_name"

request.body = AddFacesBase64Req(

external_fields="{\"timestamp\":12}",

)

response = client.add_faces_by_base64(request) print(response)

● 删除人脸

def deleteFace():

# Delete Face By FaceId try:

request = DeleteFaceByFaceIdRequest() request.face_set_name = "face_set_name"

request.face_id = "LkPJblq6"

response = client.delete_face_by_face_id(request) print(response)

# Delete Face By ExternalImageId try:

request = DeleteFaceByExternalImageIdRequest() request.face_set_name = "face_set_name"

request.external_image_id = "external_image_id"

response = client.delete_face_by_external_image_id(request) print(response)

def batchDeleteFaces():

try:

request = BatchDeleteFacesRequest() request.face_set_name = "face_set_name"

request.body = DeleteFacesBatchReq(

filter="age:[20 TO 30]"

)

response = client.batch_delete_faces(request) print(response)

● 更新人脸

def updateFace():

try:

request = UpdateFaceRequest()

request.face_set_name = "face_set_name"

request.body = UpdateFaceReq(face_id="LkPJblq6") response = client.update_face(request)

print(response)

(19)

● 查询人脸

def showFaces():

# Show Faces By FaceId try:

request = ShowFacesByFaceIdRequest() request.face_set_name = "face_set_name"

request.face_id = "LkPJblq6"

response = client.show_faces_by_face_id(request) print(response)

print(e.status_code) print(e.request_id) print(e.error_code) print(e.error_msg) # Show Faces By Limit try:

request = ShowFacesByLimitRequest() request.face_set_name = "face_set_name"

request.offset = 0 request.limit = 10

response = client.show_faces_by_limit(request) print(response)

def detectLiveByBase64():

try:

request = DetectLiveByBase64Request() request.body = LiveDetectBase64Req(

actions="1,2,3,4",

video_base64="/9j/4AAQSkZJRgABAQAAAQABAAD..."

)

response = client.detect_live_by_base64(request) print(response)

def detectLiveFaceByBase64():

try:

request = DetectLiveFaceByBase64Request() request.body = LiveDetectFaceBase64Req(

)

response = client.detect_live_face_by_base64(request) print(response)

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

(20)

(21)

4 ^{Go SDK}

本章节介绍人脸识别服务Go SDK，您可以参考本章节进行快速集成开发。

准备工作

● 已开通人脸识别服务。如未开通，请登录人脸识别管理控制台人脸识别管理控制台开通所需服务。

● 已具备开发环境，Go SDK 支持 go 1.14 及以上版本，可执行 go version 检查当前 Go 的版本信息。

（SK）。

图4-1 我的凭证

安装 SDK

使用SDK前需要安装华为云Go SDK 库。

# 安装华为云Go库

go get -u github.com/huaweicloud/huaweicloud-sdk-go-v3

# 安装依赖

go get github.com/json-iterator/go

(22)

开始使用

详细的SDK介绍请参见SDK中心、Go SDK使用指导、Go SDK使用视频。

import ( "fmt"

"github.com/huaweicloud/huaweicloud-sdk-go-v3/core/auth/basic"

// 导入v2版本sdk

frs "github.com/huaweicloud/huaweicloud-sdk-go-v3/services/frs/v2"

"github.com/huaweicloud/huaweicloud-sdk-go-v3/services/frs/v2/model"

region "github.com/huaweicloud/huaweicloud-sdk-go-v3/services/frs/v2/region"

)

func GetCredential(ak, sk string) basic.Credentials { // Init Auth Info

return basic.NewCredentialsBuilder().

WithAk(ak).

WithSk(sk).

Build() }

// # 初始化人脸识别服务的客户端，并选择服务部署区域 func GetClient(auth basic.Credentials) *frs.FrsClient { return frs.NewFrsClient(

frs.FrsClientBuilder().

WithRegion(region.CN_NORTH_4).

WithCredential(auth).

Build()) }

4. 发送请求并查看下响应

request := &model.DetectFaceByBase64Request{}

request.Body = &model.FaceDetectBase64Req{

ImageBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.DetectFaceByBase64(request) if err == nil {

fmt.Printf("%+v\n", response) } else {

fmt.Println(err) }

说明

5. 异常处理

表4-1 异常处理

一级分类一级分类说明

ServiceResponseError 服务响应异常

url.Error url异常

(23)

response, err := client.DetectFaceByBase64(request) if err == nil {

fmt.Println(err) }

SDK demo 代码解析

● 人脸检测

request := &model.DetectFaceByBase64Request{}

attributesFaceDetectBase64Req := "2"

request.Body = &model.FaceDetectBase64Req{

Attributes: &attributesFaceDetectBase64Req, ImageBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.DetectFaceByBase64(request) if err == nil {

fmt.Println(err) }

● 人脸比对

request := &model.CompareFaceByBase64Request{}

request.Body = &model.FaceCompareBase64Req{

Image1Base64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", Image2Base64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.CompareFaceByBase64(request) if err == nil {

fmt.Println(err) }

● 人脸搜索

request := &model.SearchFaceByBase64Request{}

request.FaceSetName = "face_set_name"

var listReturnFieldsbody = []string{

"timestamp", "id",

}request.Body = &model.FaceSearchBase64Req{

ReturnFields: &listReturnFieldsbody,

ImageBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.SearchFaceByBase64(request) if err == nil {

fmt.Println(err) }

● 创建人脸库

request := &model.CreateFaceSetRequest{}

request.Body = &model.CreateFaceSetReq{

FaceSetName: "face_set_name",

}response, err := client.CreateFaceSet(request) if err == nil {

fmt.Println(err) }

● 查询人脸库

request := &model.ShowFaceSetRequest{}

response, err := client.ShowFaceSet(request)

(24)

if err == nil {

fmt.Println(err) }

request := &model.ShowAllFaceSetsRequest{}

response, err := client.ShowAllFaceSets(request) if err == nil {

fmt.Println(err) }

● 删除人脸库

request := &model.DeleteFaceSetRequest{}

response, err := client.DeleteFaceSet(request) if err == nil {

fmt.Println(err) }

● 添加人脸

request := &model.AddFacesByBase64Request{}

var externalFieldsAddFacesBase64Req interface{} = "{\"timestamp\":12}"

request.Body = &model.AddFacesBase64Req{

ExternalFields: &externalFieldsAddFacesBase64Req, ImageBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.AddFacesByBase64(request) if err == nil {

fmt.Println(err) }

● 删除人脸

request := &model.DeleteFaceByExternalImageIdRequest{}

request.ExternalImageId = "external_image_id"

response, err := client.DeleteFaceByExternalImageId(request) if err == nil {

fmt.Println(err) }

request := &model.BatchDeleteFacesRequest{}

request.Body = &model.DeleteFacesBatchReq{

Filter: "age:[20 TO 30]",

}response, err := client.BatchDeleteFaces(request) if err == nil {

fmt.Println(err) }

● 更新人脸

request := &model.UpdateFaceRequest{}

externalImageIdUpdateFaceReq:= "external_image_id"

var externalFieldsUpdateFaceReq interface{} = "{\"timestamp\":12}"

request.Body = &model.UpdateFaceReq{

FaceId: "LkPJblq6",

(25)

ExternalImageId: &externalImageIdUpdateFaceReq, ExternalFields: &externalFieldsUpdateFaceReq, }response, err := client.UpdateFace(request) if err == nil {

fmt.Println(err) }

● 查询人脸

//Show Faces By FaceId

request := &model.ShowFacesByFaceIdRequest{}

request.FaceId = "LkPJblq6"

response, err := client.ShowFacesByFaceId(request) if err == nil {

fmt.Println(err) }

//detect live by base64

request := &model.DetectLiveByBase64Request{}

request.Body = &model.LiveDetectBase64Req{

Actions: "1,2,3,4",

VideoBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.DetectLiveByBase64(request) if err == nil {

fmt.Println(err) }

//detect live face by base64

request := &model.DetectLiveFaceByBase64Request{}

request.Body = &model.LiveDetectFaceBase64Req{

ImageBase64: "/9j/4AAQSkZJRgABAQAAAQABAAD...", }response, err := client.DetectLiveFaceByBase64(request) if err == nil {

fmt.Println(err) }

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

(26)

(27)

5 ^{.NET SDK}

本章节介绍.NET SDK，您可以参考本章节进行快速集成开发。

准备工作

● 已具备开发环境，.NET SDK 适用于.NET Standard 2.0 及其以上版本；C# 4.0 及其以上版本。

（SK）。

图5-1 我的凭证

安装 SDK

使用SDK前，需要安装“HuaweiCloud.SDK.Core”和“HuaweiCloud.SDK.Frs”，有两种安装方式，分别如下。

● 使用 .NET CLI 工具

dotnet add package HuaweiCloud.SDK.Core dotnet add package HuaweiCloud.SDK.Frs

● 使用 Package Manager

Install-Package HuaweiCloud.SDK.Core Install-Package HuaweiCloud.SDK.Frs

(28)

开始使用

详细的SDK介绍，使用异步客户端，配置日志请参见SDK中心、.NET SDK使用指

导、.NET SDK视频指导。

using System;

using System.Collections.Generic;

using HuaweiCloud.SDK.Core;

using HuaweiCloud.SDK.Core.Auth;

using HuaweiCloud.SDK.Frs;

using HuaweiCloud.SDK.Frs.V2;

using HuaweiCloud.SDK.Frs.V2.Model;

2. 配置客户端连接参数 – 默认配置

// 使用默认配置

var config = HttpConfig.GetDefaultConfig();

– 网络代理（可选）

// 根据需要配置网络代理

config.ProxyHost = "proxy.huaweicloud.com";

config.ProxyPort = 8080;

config.ProxyUsername = "test";

config.ProxyPassword = "test";

– 超时配置（可选）

// 默认超时时间为120秒，可根据需要调整 config.Timeout = 120;

– SSL配置（可选）

// 根据需要配置是否跳过SSL证书验证 config.IgnoreSslVerification = true;

const string ak = "<YOUR AK>";

const string sk = "<YOUR SK>";

var auth = new BasicCredentials(ak, sk);

4. 初始化客户端

– 指定云服务region方式（推荐）

// 初始化指定云服务的客户端 {Service}Client ，以初始化FRS服务的 FrsClient 为例 var client = FrsClient.NewBuilder()

.WithCredential(auth)

.WithRegion(FrsRegion.ValueOf("cn-north-4")) .WithHttpConfig(config)

.Build();

– 指定云服务endpoint方式

// 指定终端节点，以FRS服务北京四的 endpoint 为例

String endpoint = "https://face.cn-north-4.myhuaweicloud.com";

// 初始化客户端认证信息，需要填写相应 projectId，以初始化 BasicCredentials 为例 var auth = new BasicCredentials(ak, sk, projectId);

// 初始化指定云服务的客户端 {Service}Client，以初始化FRS服务的 FrsClient 为例 var client = FrsClient.NewBuilder()

.WithCredential(auth) .WithEndPoint(endpoint) .WithHttpConfig(config) .Build();

endpoint是华为云各服务应用区域和各服务的终端节点，详情请查看地区和

终端节点。

(29)

// 以调用人脸检测接口 DetectFaceByBase64 为例 var req = new DetectFaceByBase64Request {};

req.Body = new FaceDetectBase64Req() { ImageBase64 = "图片的base64编码"

};

try{

var resp = client.DetectFaceByBase64(req);

var respStatusCode = resp.HttpStatusCode;

Console.WriteLine(respStatusCode);

}

说明

6. 异常处理

表5-1 异常处理

ConnectionExcepti

SslHandShakeExce

RequestTimeoutEx

RetryOutageExcep

ServiceResponseEx

Http响应码：

[500,]。

ClientRequestExce

Http响应码：[400, 500)

try{

}catch (RequestTimeoutException requestTimeoutException) { Console.WriteLine(requestTimeoutException.ErrorMessage);

}catch (ServiceResponseException clientRequestException) { Console.WriteLine(clientRequestException.HttpStatusCode);

(30)

Console.WriteLine(clientRequestException.ErrorCode);

Console.WriteLine(clientRequestException.ErrorMsg);

}catch (ConnectionException connectionException) { Console.WriteLine(connectionException.ErrorMessage);

}

SDK demo 代码解析

● 人脸检测

var req = new DetectFaceByBase64Request {

};

req.Body = new FaceDetectBase64Req() {

ImageBase64 = "图片的base64编码"

};

try {

}

catch (RequestTimeoutException requestTimeoutException) {

Console.WriteLine(requestTimeoutException.ErrorMessage);

}

catch (ServiceResponseException clientRequestException) {

Console.WriteLine(clientRequestException.HttpStatusCode);

}

catch (ConnectionException connectionException) {

Console.WriteLine(connectionException.ErrorMessage);

}

● 人脸比对

var req = new CompareFaceByBase64Request {

};

req.Body = new FaceCompareBase64Req() {

Image1Base64 = "图片1的base64编码", Image2Base64 = "图片2的base64编码"

};

try {

var resp = client.CompareFaceByBase64(req);

}

(31)

● 人脸搜索

var req = new SearchFaceByBase64Request {

};

req.Body = new FaceSearchBase64Req() {

};

try {

var resp = client.SearchFaceByBase64(req);

}

● 创建人脸库

var req = new CreateFaceSetRequest {

};

req.Body = new CreateFaceSetReq() {

FaceSetName = "人脸库名称"

};

try {

var resp = client.CreateFaceSet(req);

}

● 查询人脸库

var req = new ShowFaceSetRequest {

};

try {

var resp = client.ShowFaceSet(req);

}

catch (RequestTimeoutException requestTimeoutException)

(32)

{

}

var req = new ShowAllFaceSetsRequest {

};

try {

var resp = client.ShowAllFaceSets(req);

}

● 删除人脸库

var req = new DeleteFaceSetRequest {

};

try {

var resp = client.DeleteFaceSet(req);

}

● 添加人脸

var req = new AddFacesByBase64Request {

};

req.Body = new AddFacesBase64Req()

(33)

{

};

try {

var resp = client.AddFacesByBase64(req);

}

● 删除人脸

var req = new DeleteFaceByFaceIdRequest {

FaceSetName = "人脸库名称", FaceId = "人脸ID"

};

try {

var resp = client.DeleteFaceByFaceId(req);

}

var req = new BatchDeleteFacesRequest {

};

req.Body = new DeleteFacesBatchReq() {

Filter = "过滤条件"

};

try {

var resp = client.BatchDeleteFaces(req);

}

catch (ServiceResponseException clientRequestException)

(34)

{

}

● 更新人脸

var req = new UpdateFaceRequest {

};

req.Body = new UpdateFaceReq() {

FaceId = "人脸库ID"

};

try {

var resp = client.UpdateFace(req);

}

} }

● 查询人脸

var req = new ShowFacesByFaceIdRequest {

FaceSetName = "人脸库名称", FaceId = "人脸ID"

};

try {

var resp = client.ShowFacesByFaceId(req);

}

var req = new DetectLiveByBase64Request {

(35)

};

req.Body = new LiveDetectBase64Req() {

Actions = "动作代码顺序列表",

VideoBase64 = "视频数据的base64编码"

};

try {

var resp = client.DetectLiveByBase64(req);

}

var req = new DetectLiveFaceByBase64Request {

};

req.Body = new LiveDetectFaceBase64Req() {

};

try {

var resp = client.DetectLiveFaceByBase64(req);

}

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

您只需要在API Explorer中修改接口参数，即可自动生成对应的代码示例。

(36)

(37)

6 Node.js SDK

本章节介绍新版Node.js SDK，您可以参考本章节进行快速集成开发。

准备工作

● 已具备开发环境，支持Node 10.16.1 及其以上版本。

（SK）。

图6-1 我的凭证

安装 SDK

使用SDK前，需要安装“@huaweicloud/huaweicloud-sdk-core”和

“@huaweicloud/huaweicloud-sdk-frs”。

推荐您使用 npm 安装 SDK。

npm install @huaweicloud/huaweicloud-sdk-core npm i @huaweicloud/huaweicloud-sdk-frs

开始使用

详细的SDK介绍请参见SDK中心、Node.js SDK使用指导、Node.js SDK视频指导。

(38)

const core = require('@huaweicloud/huaweicloud-sdk-core');

const frs = require("@huaweicloud/huaweicloud-sdk-frs");

2. 配置客户端链接参数 – 默认配置

const client = frs.FrsClient.newBuilder()

– 网络代理（可选）

// 使用代理服务器（可选）

client.withProxyAgent("http://username:[email protected]:8080")

– SSL配置（可选）

// 配置跳过服务端证书验证（可选）

process.env.NODE_TLS_REJECT_UNAUTHORIZED = "0"

配置AK、SK、project_id信息。华为云通过AK识别用户的身份，通过SK对请求数据进行签名验证，用于确保请求的机密性、完整性和请求者身份的正确性。

const credentials = new core.BasicCredentials() .withAk(ak)

.withSk(sk)

.withProjectId(project_id)

认证参数说明：

– ak、sk：访问秘钥信息，获取方法请参见准备工作。

– project_id：华为云项目ID，获取方法请参见准备工作。

4. 初始化客户端

指定云服务endpoint方式

// 指定终端节点，以 FRS 服务北京四的 endpoint 为例

const endpoint = "https://face.cn-north-4.myhuaweicloud.com";

const client = frs.FrsClient.newBuilder() .withCredential(credentials) .withEndpoint(endpoint) .build();

endpoint是华为云各服务应用区域和各服务的终端节点，详情请查看地区和终端

节点。

// 以调用人脸检测接口 DetectFaceByBase64 为例 const request = new frs.DetectFaceByBase64Request();

const body = new frs.FaceDetectBase64Req();

body.withImageBase64("图片的base64编码");

const result = client.detectFaceByBase64(request);

result.then(result => {

console.log("JSON.stringify(result)::" + JSON.stringify(result));

}).catch(ex => {

console.log("exception:" + JSON.stringify(ex));

});

说明

SDK demo 代码解析

● 人脸检测

const request = new frs.DetectFaceByBase64Request();

const body = new frs.FaceDetectBase64Req();

const result = client.detectFaceByBase64(request);

(39)

}).catch(ex => {

});

● 人脸比对

const request = new frs.CompareFaceByBase64Request();

const body = new frs.FaceCompareBase64Req();

body.withImage1Base64("图片1的base64编码");

body.withImage2Base64("图片2的base64编码");

const result = client.compareFaceByBase64(request);

}).catch(ex => {

});

● 人脸搜索

const request = new frs.SearchFaceByBase64Request();

const body = new frs.FaceSearchBase64Req();

const result = client.searchFaceByBase64(request);

}).catch(ex => {

});

● 创建人脸库

const request = new frs.CreateFaceSetRequest();

const body = new frs.CreateFaceSetReq();

body.withFaceSetName("人脸库名称");

const result = client.createFaceSet(request);

}).catch(ex => {

});

● 查询人脸库

const request = new frs.ShowFaceSetRequest();

request.faceSetName = "人脸库名称";

const result = client.showFaceSet(request);

}).catch(ex => {

});

const request = new frs.ShowAllFaceSetsRequest();

const result = client.showAllFaceSets(request);

}).catch(ex => {

});

● 删除人脸库

const request = new frs.DeleteFaceSetRequest();

request.faceSetName = "人脸库名称";

const result = client.deleteFaceSet(request);

}).catch(ex => {

});

CSharp SDK_人脸识别服务 FRS_SDK参考_SDK（旧版）_华为云

SDK 参考

华为技术有限公司

https://www.huawei.com

[email protected]

目 录

1 简介...1

2 Java SDK...3

3 Python SDK... 11

4 Go SDK... 18

5 .NET SDK...24

6 Node.js SDK...34

7 PHP SDK...39

8 C++ SDK... 47

A 修订记录... 57

1 简介

接口与 API 对应关系

2 Java SDK

准备工作

安装 SDK

Maven，安装完成后您只需要在Java项目的pom.xml文件中加入相应的依赖项即可。

开始使用

Java SDK使用视频。

SDK demo 代码解析

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

3 Python SDK

准备工作

安装 SDK

开始使用

导、Python SDK使用视频。

SDK 代码解析

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

4 Go SDK

准备工作

安装 SDK

开始使用

SDK demo 代码解析

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

5 .NET SDK

准备工作

安装 SDK

开始使用

导、.NET SDK视频指导。

终端节点 。

SDK demo 代码解析

SDK 代码自动生成

API Explorer提供API检索及平台调试，支持全量快速检索、可视化调试、帮助文档查

6 Node.js SDK

准备工作

安装 SDK

开始使用

节点 。

SDK demo 代码解析

目录

1 ^简介

2 ^{Java SDK}

4 ^{Go SDK}

5 ^{.NET SDK}

终端节点。

节点。