JAVA实现人脸识别，活体检测之百度API

----------------------------------------------------------------

GitHub：https://github.com/reamZMX/led-

-----------------------8/16更新---------------------------

有人问源码:

链接：https://pan.baidu.com/s/1WP37IBacu6VZwtNOUDbSYg 密码：vdzm

在这

--------------------------------------------------

吃水不忘挖井人，挖井人博客在此：挖井人一，挖井人二；

然后是官方的使用说明文档：地址；

以上，感谢。

首先引入jar：

maven:

    <dependency><groupId>com.baidu.aip</groupId><artifactId>java-sdk</artifactId><version>4.3.2</version></dependency>

然后获取所需的AppID、APIKey、SecretKey：

进入百度云，创建应用：

创建完成之后可看见如下：

获取AppID、APIKey、SecretKey后，导入一下代码：

public class FaceSpot {private static final String AppID = "";private static final String APIKey = "";private static final String SecretKey = "";static AipFace client = null;static {client = new AipFace(AppID, APIKey, SecretKey);// 可选：设置网络连接参数client.setConnectionTimeoutInMillis(2000);client.setSocketTimeoutInMillis(60000);}public static void main(String[] args) throws IOException {String filePath = "F:/3.jpg";byte[] imgData = FileToByte(new File(filePath));System.out.println(detectFace(imgData,"1"));
//      String filePath1 = "F:/3.jpg";
//      String filePath2 = "F:/7.jpg";
//      byte[] imgData1 = FileUtil.readFileByBytes(filePath1);
//      byte[] imgData2 = FileUtil.readFileByBytes(filePath2);
//      System.out.println(faceverify(imgData1));}/*** 人脸检测* * @return* @throws IOException*/public static String detectFace(File file, String max_face_num) {try {return detectFace(FileToByte(file), max_face_num);} catch (IOException e) {// TODO Auto-generated catch blocke.printStackTrace();}return null;}/*** 人脸检测* * @return* @throws IOException*/public static String detectFace(byte[] arg0, String max_face_num) {try {HashMap<String, String> options = new HashMap<String, String>();options.put("face_field", "age,beauty,expression,faceshape,gender,glasses,race,qualities");options.put("max_face_num", "2");options.put("face_type", "LIVE");// 图片数据String imgStr = Base64Util.encode(arg0);String imageType = "BASE64";JSONObject res = client.detect(imgStr, imageType, options);System.out.println(res.toString(2));return res.toString();} catch (Exception e) {e.printStackTrace();}return null;}/*** 人脸比对* @param file1* @param file2* @return*/public static String matchFace(File file1, File file2) {try {return matchFace(FileToByte(file1), FileToByte(file2));} catch (IOException e) {// TODO Auto-generated catch blocke.printStackTrace();}return null;}/*** 人脸比对* * @param arg0*            人脸1* @param arg1*            人脸2* @return*/public static String matchFace(byte[] arg0, byte[] arg1) {String imgStr1 = Base64Util.encode(arg0);String imgStr2 = Base64Util.encode(arg1);MatchRequest req1 = new MatchRequest(imgStr1, "BASE64");MatchRequest req2 = new MatchRequest(imgStr2, "BASE64");ArrayList<MatchRequest> requests = new ArrayList<MatchRequest>();requests.add(req1);requests.add(req2);JSONObject res = client.match(requests);return res.toString();}/*** 人脸搜索* @param file* @param groupIdList* @param userId* @return*/public static String searchFace(File file, String groupIdList, String userId) {try {return searchFace(FileToByte(file), groupIdList, userId);} catch (IOException e) {// TODO Auto-generated catch blocke.printStackTrace();}return null;}/*** 人脸搜索* * @param arg0* @param groupIdList* @return*/public static String searchFace(byte[] arg0, String groupIdList, String userId) {String imgStr = Base64Util.encode(arg0);String imageType = "BASE64";HashMap<String, String> options = new HashMap<String, String>();options.put("quality_control", "NORMAL");options.put("liveness_control", "LOW");if (userId != null) {options.put("user_id", userId);}options.put("max_user_num", "1");JSONObject res = client.search(imgStr, imageType, groupIdList, options);return res.toString(2);}/*** 增加用户* @param file* @param userInfo* @param userId* @param groupId* @return*/public static String addUser(File file, String userInfo, String userId, String groupId) {try {return addUser(FileToByte(file), userInfo, userId, groupId);} catch (IOException e) {// TODO Auto-generated catch blocke.printStackTrace();}return null;}/*** 增加用户* * @param arg0* @param userInfo* @param userId* @param groupId* @return*/public static String addUser(byte[] arg0, String userInfo, String userId, String groupId) {String imgStr = Base64Util.encode(arg0);String imageType = "BASE64";HashMap<String, String> options = new HashMap<String, String>();options.put("user_info", userInfo);options.put("quality_control", "NORMAL");options.put("liveness_control", "LOW");JSONObject res = client.addUser(imgStr, imageType, groupId, userId, options);return res.toString(2);}public static String updateUser(File file, String userInfo, String userId, String groupId) {try {return updateUser(FileToByte(file), userInfo, userId, groupId);} catch (IOException e) {// TODO Auto-generated catch blocke.printStackTrace();}return null;}/*** 更新用户* * @param arg0* @param userInfo* @param userId* @param groupId* @return*/public static String updateUser(byte[] arg0, String userInfo, String userId, String groupId) {String imgStr = Base64Util.encode(arg0);String imageType = "BASE64";HashMap<String, String> options = new HashMap<String, String>();if (userInfo != null) {options.put("user_info", userInfo);}options.put("quality_control", "NORMAL");options.put("liveness_control", "LOW");JSONObject res = client.updateUser(imgStr, imageType, groupId, userId, options);return res.toString(2);}/*** 删除用户人脸信息* @param userId* @param groupId* @param faceToken* @return*/public static String deleteUserFace(String userId, String groupId, String faceToken) {HashMap<String, String> options = new HashMap<String, String>();// 人脸删除JSONObject res = client.faceDelete(userId, groupId, faceToken, options);return res.toString();}/*** 查询用户信息* @param userId* @param groupId* @return*/public static String searchUserInfo(String userId, String groupId) {HashMap<String, String> options = new HashMap<String, String>();// 用户信息查询JSONObject res = client.getUser(userId, groupId, options);return res.toString(2);}/*** 获取用户人脸列表* @param userId* @param groupId* @return*/public static String getUserFaceList(String userId, String groupId) {HashMap<String, String> options = new HashMap<String, String>();// 获取用户人脸列表JSONObject res = client.faceGetlist(userId, groupId, options);return res.toString(2);}/*** 获取一组用户* @param groupId* @param returnNum* @return*/public static String getGroupUsers(String groupId, String returnNum) {HashMap<String, String> options = new HashMap<String, String>();options.put("start", "0");if (returnNum != null) {options.put("length", returnNum);}// 获取用户列表JSONObject res = client.getGroupUsers(groupId, options);return res.toString(2);}/*** 组用户复制* @param userId* @param srcGroupId* @param dstGroupId* @return*/public static String userCopy(String userId, String srcGroupId, String dstGroupId) {HashMap<String, String> options = new HashMap<String, String>();options.put("src_group_id", srcGroupId);options.put("dst_group_id", dstGroupId);// 复制用户JSONObject res = client.userCopy(userId, options);return res.toString(2);}/*** 删除用户* @param userId* @param groupId* @return*/public static String deleteUser(String userId, String groupId) {HashMap<String, String> options = new HashMap<String, String>();// 人脸删除JSONObject res = client.deleteUser(groupId, userId, options);return res.toString();}/*** 增加组信息* @param groupId* @return*/public static String addGroup(String groupId) {HashMap<String, String> options = new HashMap<String, String>();// 创建用户组JSONObject res = client.groupAdd(groupId, options);return res.toString();}/*** 删除* @param groupId* @return*/public static String deleteGroup(String groupId) {HashMap<String, String> options = new HashMap<String, String>();// 创建用户组JSONObject res = client.groupDelete(groupId, options);return res.toString();}/*** 获取组列表* @param length* @return*/public static String getGroupList(String length) {HashMap<String, String> options = new HashMap<String, String>();options.put("start", "0");options.put("length", length);// 组列表查询JSONObject res = client.getGroupList(options);return res.toString();}/*** 活体检测* @param arg0* @return*/public static String faceverify(byte[] arg0){String imgStr = Base64Util.encode(arg0);String imageType = "BASE64";FaceVerifyRequest req = new FaceVerifyRequest(imgStr, imageType);ArrayList<FaceVerifyRequest> list = new ArrayList<FaceVerifyRequest>();list.add(req);JSONObject res = client.faceverify(list);return res.toString();}private static byte[] FileToByte(File file) throws IOException {// 将数据转为流InputStream content = new FileInputStream(file);ByteArrayOutputStream swapStream = new ByteArrayOutputStream();byte[] buff = new byte[100];int rc = 0;while ((rc = content.read(buff, 0, 100)) > 0) {swapStream.write(buff, 0, rc);}// 获得二进制数组return swapStream.toByteArray();}}

填入自己的AppID, APIKey, SecretKey；

函数的参数可自行根据官方文档自己改一改，文件读取那边提供一下小帅的FileUtil并且处理好了异常：

public class FileUtil {/*** 根据文件路径读取byte[] 数组*/public static byte[] readFileByBytes(String filePath) throws IOException {File file = new File(filePath);if (!file.exists()) {throw new FileNotFoundException(filePath);} else {ByteArrayOutputStream bos = new ByteArrayOutputStream((int) file.length());BufferedInputStream in = null;try {in = new BufferedInputStream(new FileInputStream(file));short bufSize = 1024;byte[] buffer = new byte[bufSize];int len1;while (-1 != (len1 = in.read(buffer, 0, bufSize))) {bos.write(buffer, 0, len1);}byte[] var7 = bos.toByteArray();return var7;} finally {try {if (in != null) {in.close();}} catch (IOException var14) {var14.printStackTrace();}bos.close();}}}
}

然后就是测试啦，测试的话，随便百度了一张美女的照片，然后看返回：

{"timestamp": 1528422299,"result": {"face_list": [{"expression": {"probability": 0.9997190833,"type": "smile"},"face_probability": 1,"glasses": {"probability": 0.9999995232,"type": "none"},"location": {"height": 102,"rotation": -2,"width": 109,"left": 102.9398575,"top": 56.39219284},"age": 19,"gender": {"probability": 0.9999992847,"type": "female"},"face_shape": {"probability": 0.7517765164,"type": "heart"},"face_token": "706eaee3240ef0cc679ab209b1b71e0d","beauty": 69.47167206,"race": {"probability": 0.9999864101,"type": "yellow"},"angle": {"yaw": -2.858289719,"roll": -2.302281618,"pitch": 9.867022514}}],"face_num": 1},"cached": 0,"error_code": 0,"log_id": 3671200838,"error_msg": "SUCCESS"
}

嗯，确实是美女，

"beauty": 69.47167206

好，然后通过浏览器开启摄像头来实时的试下，先上效果图：

好，大家不用在意这些细节，比如颜值啥的。。

我觉得颜值这项可能是假的，因为我做搞怪的表情时分数反而高一些？？？？？？？？？？

上代码吧。。

首先是Controller：

@Controller
@RequestMapping(value = "/faceRecognition")
public class faceRecognitionController {/*** 人脸检测测试页面* @return* @throws Exception  */@RequestMapping(value = "/test.do")public ModelAndView queryVoi() throws Exception {ModelAndView modelAndView = new ModelAndView();modelAndView.setViewName("/artificialIntelligence/faceRecognition/test");return modelAndView;}/*** 请求人脸检测* @return* @throws Exception  */@RequestMapping(value = "/save.do")@ResponseBodypublic Map<String, Object> queryService(@RequestParam("the_file") MultipartFile file) {Map<String, Object> modelMap = new HashMap<String, Object>();try {//将数据转为流InputStream content = file.getInputStream();ByteArrayOutputStream swapStream = new ByteArrayOutputStream();  byte[] buff = new byte[100];  int rc = 0;  while ((rc = content.read(buff, 0, 100)) > 0) {  swapStream.write(buff, 0, rc);  }  //获得二进制数组byte[] in2b = swapStream.toByteArray(); //调用人脸检测的方法String  str = FaceSpot.detectFace(in2b,""+1);JSONObject job = new JSONObject(FaceSpot.faceverify(in2b));System.out.println(job.toString());JSONObject testData = job.getJSONObject("result");//System.out.println(testData.get("face_liveness"));JSON json = JSON.parseObject(str);FaceV3DetectBean bean = JSON.toJavaObject(json, FaceV3DetectBean.class);JSONArray arr = new JSONArray();for(int i=0;i<bean.getResult().getFace_list().size();i++){JSONObject jsonObject = new JSONObject();//获取年龄int ageOne = bean.getResult().getFace_list().get(i).getAge();//处理年龄String age =String.valueOf(new BigDecimal(ageOne).setScale(0, BigDecimal.ROUND_HALF_UP));jsonObject.put("age", age);//获取美丑打分Double beautyOne = (Double) bean.getResult().getFace_list().get(i).getBeauty();//处理美丑打分String beauty =String.valueOf(new BigDecimal(beautyOne).setScale(0, BigDecimal.ROUND_HALF_UP));jsonObject.put("beauty", beauty);//获取性别  male(男)、female(女)String gender = (String) bean.getResult().getFace_list().get(i).getGender().getType();jsonObject.put("gender", gender);//获取是否带眼睛 0-无眼镜，1-普通眼镜，2-墨镜String glasses =  bean.getResult().getFace_list().get(i).getGlasses().getType();jsonObject.put("glasses", String.valueOf(glasses));//获取是否微笑，0，不笑；1，微笑；2，大笑String expression =  bean.getResult().getFace_list().get(i).getExpression().getType();jsonObject.put("expression", String.valueOf(expression));arr.put(jsonObject);}modelMap.put("strjson", arr.toString());modelMap.put("face_liveness", testData.get("face_liveness"));modelMap.put("success", true);} catch (Exception e) {e.printStackTrace();modelMap.put("success", false);modelMap.put("data", e.getMessage());}return modelMap;}}

用的是fastjson，自行加jar包。

然后是jsp：

<%@ page language="java" contentType="text/html; charset=UTF-8" pageEncoding="UTF-8"%>
<%@ taglib prefix="c" uri="http://java.sun.com/jsp/jstl/core"%>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1">
<title>人脸检测</title>
<script src="js/jquery-2.1.1.js" type="text/javascript" charset="utf-8"></script>
<script>//判断浏览器是否支持HTML5 Canvaswindow.onload = function () {try {//动态创建一个canvas元 ，并获取他2Dcontext。如果出现异常则表示不支持 document.createElement("canvas").getContext("2d");document.getElementById("support").innerHTML = "浏览器支持HTML5 CANVAS";}catch (e) {document.getElementByIdx("support").innerHTML = "浏览器不支持HTML5 CANVAS";}};//这段代 主要是获取摄像头的视频流并显示在Video 签中window.addEventListener("DOMContentLoaded", function () {var canvas = document.getElementById("canvas"),context = canvas.getContext("2d"),video = document.getElementById("video"),videoObj = { "video": true },errBack = function (error) {console.log("Video capture error: ", error.code);};//拍照按钮
//      $("#snap").click(function () {
//          context.drawImage(video, 0, 0, 330, 250);
//          })//拍照每秒一次setInterval(function(){context.drawImage(video, 0, 0, 330, 250)CatchCode();},1000);//navigator.getUserMedia这个写法在Opera中好像是navigator.getUserMedianow//更新兼容火狐浏览器if (navigator.getUserMedia || navigator.webkitGetUserMedia || navigator.mozGetUserMedia) {navigator.getUserMedia=navigator.getUserMedia || navigator.webkitGetUserMedia || navigator.mozGetUserMedia;navigator.getUserMedia(videoObj, function (stream) {video.srcObject  = stream;video.play();}, errBack);}}, false);function dataURItoBlob(base64Data) {var byteString;if (base64Data.split(',')[0].indexOf('base64') >= 0)byteString = atob(base64Data.split(',')[1]);elsebyteString = unescape(base64Data.split(',')[1]);var mimeString = base64Data.split(',')[0].split(':')[1].split(';')[0];var ia = new Uint8Array(byteString.length);for (var i = 0; i < byteString.length; i++) {ia[i] = byteString.charCodeAt(i);}return new Blob([ia], {type:mimeString});}//上传服务器function CatchCode() {var canvans = document.getElementById("canvas");//获取浏览器页面的画布对象//以下开始编 数据var imageBase64 = canvans.toDataURL();var blob = dataURItoBlob(imageBase64);  // 上一步中的函数var fd = new FormData(document.forms[0]);fd.append("the_file", blob, 'image.png');//将图像转换为base64数据$.ajax({type:"POST",url:"faceRecognition/save.do",processData: false,     // 必须contentType: false,     // 必须data:fd,datatype: "json",success:function(data){var mes = eval(data);//alert(mes.success);//var jsonObj =   $.parseJSON(mes.strjson); //alert(jsonObj[0].age);if (mes.success) {//alert(mes.strjson);var jsonObj =  $.parseJSON(mes.strjson); //alert(jsonObj);var age = jsonObj[0].age;var beauty = jsonObj[0].beauty;var gendergender = jsonObj[0].gender;var glasses = jsonObj[0].glasses;var expression = jsonObj[0].expression$("#age").html(age);$("#beauty").html(beauty);$("#faceverify").html(mes.face_liveness);if(gendergender == 'male'){$("#gendergender").html("男");}else{$("#gendergender").html("女");}if(glasses == 'none'){$("#glasses").html("未戴眼镜");}else if(glasses == 'common'){$("#glasses").html("戴了普通眼镜");}else{$("#glasses").html("戴了墨镜");}if(expression == 'none'){$("#expression").html("不笑");}else if(expression == 'smile'){$("#expression").html("微笑");}else{$("#expression").html("大笑");}}},error: function(){//请求出错处理alert("出情况了");}         });}
</script>
<style>
.div-a{ float:left;width:60%;height:60%;border:1px solid #F00}
.div-b{ float:left;width:39%;height:60%;border:1px solid #000}
span{ font-size:25px }
</style>
</head>
<body> <!-- 左边区域 --><div class="div-a" id="contentHolder"><video id="video" width="100%" height="60%" autoplay></video><canvas style="" hidden="hidden"  id="canvas" width="520" height="250"></canvas></div> <!-- 右边区域 --><div class="div-b" ><!-- 测试按钮 -->
<!--         <input type="button" id="snap" style="width:100px;height:35px;" value="拍 照" /> -->
<!--         <input type="button" onclick="CatchCode();" style="width:100px;height:35px;" value="上传服务器" /> --><h1>人脸检测实时数据</h1><span>年龄：</span><span id="age"></span><br/><span>颜值：</span><span id="beauty" ></span><br/><span>性别：</span><span id="gendergender"></span><br/><span>是否戴眼镜:</span><span id="glasses"></span><br/><span>表情：</span><span id="expression"></span><br/><span>活体分数：</span><span id="faceverify"></span><br/></div> </body>
</html>

这个貌似只支持火狐？？我用其他浏览器摄像头并未开启，有哪位大神知道代码还请告诉在下，谢谢指教。

这边有个活体指数，我测试了下，貌似还挺准，我首先百度了一张胡歌的照片放在摄像头上，emmmm.....颜值80，但是活体指数就很低，然后本人上，活体指数一直在0.95之上浮动，还是可靠的。

好，告辞！

JAVA实现人脸识别，活体检测之百度API相关推荐

人脸识别活体检测之张张嘴和眨眨眼
暑 [这段时间有点忙,终于截止今天2018.06.22完成了人脸识别的最后一道程序--活体检测之眨眨眼和张张嘴] 关于人脸识别的内容我之前也写过好几篇博文,其中有: {java实现人脸识别源码} ...
dlib实现人脸识别+活体检测
目录: 一:dlib的shape_predictor_68_face_landmarks模型二.眨眼检测三.张口检测四.眨眼检测+张口检测五.人脸识别六.人脸识别+活体检测七.人脸识别破解 ...
人脸识别--活体检测(眨眼检测)
人脸识别在现在很多的项目中都有应用,最常见的就属此次315曝光的支付宝刷脸登录,当然支付宝也出来做了澄清,我们还是需要相信科技的.支付宝的刷脸可谓是相当成熟了.下面我们来简单的分析一下支付宝的刷脸登录 ...
Android开发之虹软人脸识别活体检测基本步骤
首先,我简单说下虹软的人脸识别基本步骤: 1.你的设置好设置视频模式方向用于人脸检测有如下几个可设置方向 //设置视频模式全方向人脸检测ConfigUtil.setFtOrient(this, Fa ...
微信公众号人脸识别|活体检测SDK|人证比对API
一.概述人脸识别特指利用分析比较人脸视觉特征信息进行身份鉴别的计算机技术.人脸识别是一项热门的计算机技术研究领域,可以将人脸明暗侦测,自动调整动态曝光补偿,人脸追踪侦测,自动调整影像放大:它属于生物 ...
人脸识别活体检测之张张嘴和眨眨眼——Json_Parsing
/** * @Title: Json_Parsing.java * @Package org.util * @Description: TODO该方法的主要作用: * @author A18ccms ...
人脸识别活体检测之张张嘴和眨眨眼——readme
人脸检测返回参数详情 ----------------------------------------------------------------------------------------- ...
人脸识别活体检测之张张嘴和眨眨眼——Landmark
/** * @Title: Landmark.java * @Package org.entity * @Description: TODO该方法的主要作用: * @author A18ccms A1 ...
Android开发之虹软人脸识别活体检测SDK包Bitmap转NV21方法
/** * Bitmap 转化为 ARGB 数据,再转化为 NV21 数据 * * @param src 传入的 Bitmap,格式为 Bitmap.Config.ARGB_8888 * @param ...

JAVA实现人脸识别，活体检测之百度API

JAVA实现人脸识别，活体检测之百度API相关推荐

最新文章

热门文章