本篇文章将介绍如何使用Python利用OpenCV图像捕捉,配合强大的Mediapipe库来实现手势动作检测与识别;将识别结果实时同步至Unity中,实现手势模型在Unity中运动身体结构识别
Demo展示:https://hackathon2022.juejin.cn/#/works/detail?unique=WJoYomLPg0JOYs8GazDVrw
本篇文章所用的技术会整理后开源,后续可以持续关注:
GitHub:https://github.com/BIGBOSS-dedsec
项目下载地址:https://github.com/BIGBOSS-dedsec/OpenCV-Unity-To-Build-3DHands
CSDN: https://blog.csdn.net/weixin_50679163?type=edu
同时本篇文章实现的技术参加了稀土掘金2022编程挑战赛-游戏赛道
作品展示:https://hackathon2022.juejin.cn/#/works/detail?unique=WJoYomLPg0JOYs8GazDVrw
项目的实现,核心是强大的Mediapipe ,它是google的一个开源项目:
功能 | 详细 |
---|---|
人脸检测 FaceMesh | 从图像/视频中重建出人脸的3D Mesh |
人像分离 | 从图像/视频中把人分离出来 |
手势跟踪 | 21个关键点的3D坐标 |
人体3D识别 | 33个关键点的3D坐标 |
物体颜色识别 | 可以把头发检测出来,并图上颜色 |
Mediapipe Dev
以上是Mediapipe的几个常用功能 ,这几个功能我们会在后续一一讲解实现
Python安装Mediapipe
pip install mediapipe==0.8.9.1
也可以用 setup.py 安装
https://github.com/google/mediapipe
Python 3.7
Mediapipe 0.8.9.1
Numpy 1.21.6
OpenCV-Python 4.5.5.64
OpenCV-contrib-Python 4.5.5.64
实测也支持Python3.8-3.9
本项目在Unity中实现实时动作捕捉的核心是通过本地UDP与socket 进行通信
关于数据文件部分,详细可以查看OpenCV+Mediapipe人物动作捕捉与Unity引擎的结合中对数据文件部分的讲解和使用
摄像头捕捉部分:
import cv2
cap = cv2.VideoCapture(0) #OpenCV摄像头调用:0=内置摄像头(笔记本) 1=USB摄像头-1 2=USB摄像头-2
while True:
success, img = cap.read()
imgRGB = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) #cv2图像初始化
cv2.imshow("HandsImage", img) #CV2窗体
cv2.waitKey(1) #关闭窗体
视频帧率计算
import time
#帧率时间计算
pTime = 0
cTime = 0
while True
cTime = time.time()
fps = 1 / (cTime - pTime)
pTime = cTime
cv2.putText(img, str(int(fps)), (10, 70), cv2.FONT_HERSHEY_PLAIN, 3,
(255, 0, 255), 3) #FPS的字号,颜色等设置
Socket通信:
定义Localhost和post端口地址
sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
serverAddressPort = ("127.0.0.1", 5052) # 定义IP和端口
# 发送数据
sock.sendto(str.encode(str(data)), serverAddressPort)
手势动作捕捉:
if hands:
# Hand 1
hand = hands[0]
lmList = hand["lmList"]
for lm in lmList:
data.extend([lm[0], h - lm[1], lm[2]])
from cvzone.HandTrackingModule import HandDetector
import cv2
import socket
cap = cv2.VideoCapture(0)
cap.set(3, 1280)
cap.set(4, 720)
success, img = cap.read()
h, w, _ = img.shape
detector = HandDetector(detectionCon=0.8, maxHands=2)
sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
serverAddressPort = ("127.0.0.1", 5052)
while True:
success, img = cap.read()
hands, img = detector.findHands(img)
data = []
if hands:
# Hand 1
hand = hands[0]
lmList = hand["lmList"]
for lm in lmList:
data.extend([lm[0], h - lm[1], lm[2]])
sock.sendto(str.encode(str(data)), serverAddressPort)
cv2.imshow("Image", img)
cv2.waitKey(1)
在Unity中,我们需要搭建一个人物的模型,这里需要一个21个Sphere作为手势的特征点和21个Cube作为中间的支架
具体文件目录如下:
Line的编号对应手势模型特征点
本代码功能将Socket发送的数据进行接收
using UnityEngine;
using System;
using System.Text;
using System.Net;
using System.Net.Sockets;
using System.Threading;
public class UDPReceive : MonoBehaviour
{
Thread receiveThread;
UdpClient client;
public int port = 5052;
public bool startRecieving = true;
public bool printToConsole = false;
public string data;
public void Start()
{
receiveThread = new Thread(
new ThreadStart(ReceiveData));
receiveThread.IsBackground = true;
receiveThread.Start();
}
private void ReceiveData()
{
client = new UdpClient(port);
while (startRecieving)
{
try
{
IPEndPoint anyIP = new IPEndPoint(IPAddress.Any, 0);
byte[] dataByte = client.Receive(ref anyIP);
data = Encoding.UTF8.GetString(dataByte);
if (printToConsole) { print(data); }
}
catch (Exception err)
{
print(err.ToString());
}
}
}
}
这里是每个Line对应cs文件,实现功能:使特征点和Line连接在一起
using System.Collections;
using System.Collections.Generic;
using UnityEngine;
public class LineCode : MonoBehaviour
{
LineRenderer lineRenderer;
public Transform origin;
public Transform destination;
void Start()
{
lineRenderer = GetComponent<LineRenderer>();
lineRenderer.startWidth = 0.1f;
lineRenderer.endWidth = 0.1f;
}
// 连接两个点
void Update()
{
lineRenderer.SetPosition(0, origin.position);
lineRenderer.SetPosition(1, destination.position);
}
}
这里是读取上文识别并保存的手势动作数据,并将每个子数据循环遍历到每个Sphere点,使特征点随着摄像头的捕捉进行运动
using System.Collections;
using System.Collections.Generic;
using UnityEngine;
public class HandTracking : MonoBehaviour
{
public UDPReceive udpReceive;
public GameObject[] handPoints;
void Start()
{
}
void Update()
{
string data = udpReceive.data;
data = data.Remove(0, 1);
data = data.Remove(data.Length - 1, 1);
print(data);
string[] points = data.Split(',');
print(points[0]);
for (int i = 0; i < 21; i++)
{
float x = 7-float.Parse(points[i * 3]) / 100;
float y = float.Parse(points[i * 3 + 1]) / 100;
float z = float.Parse(points[i * 3 + 2]) / 100;
handPoints[i].transform.localPosition = new Vector3(x, y, z);
}
}
}
Good Luck,Have Fun and Happy Coding!!!