0、任务需求

采集摄像头、烟雾传感器、温湿度传感器 （DHT11）、火焰传感器、光敏传感器

执行器：风扇、花洒、led灯

pc端订阅：video_stream、shu_data pc端发布：pc_data

树莓派：发布视频流到：video_stream， 发布数据：shu_data，接收(订阅)数据在:pc_data

温度：tem 温度阈值: th_tem

湿度：hum 湿度阈值:th_hum

空气质量:mq2 空气阈值: th_mq2

光照：light 光照阈值: th_light

火焰: fire

0、 树莓派的基本操作

1、如何查看树莓派的引脚资源

pinout

2、固定CPU的频率

树莓派4B全核心最大频率为1.5 GHz，单核最大频率为1.5 GHz

所以可以设置cpu的频率为1.5GHz

方法一

用cpufrequtils工具

sudo apt install cpufrequtils   #安装该工具

alias get_cpu='sudo cpufreq-info -w -m'             #在.bashrc中定义快捷键，用于查看cpu频率
alias set_cpu='sudo cpufreq-set -f 1.5GHz'          #在.bashrc中定义快捷键，用于设置cpu频率（固定频率）
alias reset_cpu='sudo cpufreq-set -g ondemand'      #在.bashrc中定义快捷键，用于恢复cpu频率调节的默认方式（ondemand)

方法二（推荐）

修改/boot/config.txt文件

sudo nano /boot/config.txt
# 放开修改 arm_freq=1500

查看是否成功

get_cpu

3、设置一个树莓派启动的时候自己设置环境的脚本

sudo nano /etc/rc.local

# 在rc.local的exit 0前面设置   /home/xq/startup.sh启动脚本的位置
/bin/bash /home/xq/startup.sh &

1、qt编译mqtt

2、树莓派读取dht11

1、利用Adafruit_DHT

安装dht11驱动库

pip install Adafruit_DHT

import Adafruit_DHT
import time
sensor = Adafruit_DHT.DHT11

pin = 4  # 设置GPIO引脚号

while True:
    humidity, temperature = Adafruit_DHT.read_retry(sensor, pin)
    if humidity is not None and temperature is not None:
        print('温度：{0:0.1f}°C  湿度：{1:0.1f}%'.format(temperature, humidity))
    else:
        print('读取失败，请检查传感器连接是否正确。')
        time.sleep(2)

2、到时候补充，我自己试的时候一直读不出数据

3、使用usb摄像头在树莓派上做人脸识别

关于树莓派的摄像头如何使用我在 树莓派使用摄像头网络实时监控 | 哈哈 (hahaxiong0204.github.io)这个博客上已经说明好了。

并且在树莓派上必须安装好opencv，numpy

sudo apt-get install python3-opencv -y
sudo apt-get install numpy

以下代码是通过mqtt向上位机传输数据的，也可以在树莓派上显示，打开对应的代码：cv2.imshow('image', img)

在文件路径下必须要有haarcascade_frontalface_alt2.xml文件

# face_my.py

import cv2
import os
import numpy as np
import pickle
import paho.mqtt.client as mqtt
import base64
import time

# MQTT连接参数
mqtt_broker = "192.168.137.1"  # 请替换为PC的IP地址
mqtt_port = 1883

mqtt_topic = "video_stream"

# 创建MQTT客户端
client = mqtt.Client("RaspberryPi")
client.connect(mqtt_broker, mqtt_port, 60)

cap = cv2.VideoCapture(0)
cap.set(3, 640)  # set video width
cap.set(4, 480)  # set video height

face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_alt2.xml')
current_id = 0
label_ids = {}
x_train = []
y_labels = []

def face_recognition():
    num_count = 0
    label_id_name = {}
    with open("label.pickle", 'rb') as f:
        origin_labels = pickle.load(f).items()
        # print("源文件数据:" + origin_labels)
        label_id_name = {v: h for h, v in origin_labels}
    print(label_id_name)
    recognizer = cv2.face.LBPHFaceRecognizer_create()
    recognizer.read('my_trainer.xml')
    while True:
        ret, img = cap.read();
        img = cv2.flip(img, 1)  # flip video image vertically

        gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.3, 5)
        num_count += 1
        for (x, y, w, h) in faces:
            if num_count >= 5:
                id_, conf = recognizer.predict(gray[y:y + h, x:x + w])
                print("id = ",label_id_name[id_],"conf = ",conf)
                if 20 <= conf <= 54:
                    cv2.putText(img, str(label_id_name[id_]), (x + 5, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 1,
                                (255, 255, 255), 2)
                else:
                    cv2.putText(img, str("NULL"), (x + 5, y - 5), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2)
            cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2)
            # cv2.imshow('image', img)    

        # ============================== 发送mqtt数据帧 ============================
        encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 90]
        result, encoded_frame = cv2.imencode('.jpg', img, encode_param)

        if result:
            # 将MJPEG格式的帧转换为base64编码
            frame_base64 = base64.b64encode(encoded_frame.tobytes())
            client.publish(mqtt_topic, frame_base64)
            # time.sleep(1 / 5)
        # ======================================================================
        k = cv2.waitKey(100) & 0xff  # Press 'ESC' for exiting video
        if k == 27:
            break

def input_face(name):
    count = 0
    all_num = 60
    folder = os.path.exists('./dataset/' + name)
    if not folder:  # 判断是否存在文件夹如果不存在则创建为文件夹
        os.makedirs('./dataset/' + name)  # makedirs 创建文件时如果路径不存在会创建这个路径
        print("创建文件夹成功")
    while True:
        ret, img = cap.read()
        img = cv2.flip(img, 1)  # flip video image vertically
        gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.3, 5)

        # 创建文件夹
        folder = os.path.exists('./dataset/' + name)
        if not folder:  # 判断是否存在文件夹如果不存在则创建为文件夹
            os.makedirs('./dataset/' + name)  # makedirs 创建文件时如果路径不存在会创建这个路径
            print("创建文件夹成功")
        for (x, y, w, h) in faces:
            if count%20==0:
                print("请输入任意键继续")
            if count == 0:
                print("请摆正你的脸，等待开启")
            cv2.rectangle(img, (x, y), (x + w, y + h), (255, 0, 0), 2)
            count += 1
            path_jpg = "./dataset/" + name + '/' + str(count) + ".jpg"
            cv2.imwrite(path_jpg, gray[y:y + h, x:x + w])
            # cv2.imwrite(path_jpg, img)
            print(path_jpg)
            # cv2.imshow('image', img)
        # ============================== 发送mqtt数据帧 ============================
        encode_param = [int(cv2.IMWRITE_JPEG_QUALITY), 90]
        result, encoded_frame = cv2.imencode('.jpg', img, encode_param)

        if result:
                # 将MJPEG格式的帧转换为base64编码
            frame_base64 = base64.b64encode(encoded_frame.tobytes())
            client.publish(mqtt_topic, frame_base64)
            # time.sleep(1 / 5)
        # ======================================================================

        k = cv2.waitKey(100) & 0xff  # Press 'ESC' for exiting video
        if k == 27:
            break
        elif count >= all_num:  # Take 30 face sample and stop video
            break

# 训练人脸
def train_face():
    global current_id
    recognizer = cv2.face.LBPHFaceRecognizer_create()
    for root, dirs, files in os.walk('dataset'):
        # print(root, dirs, files)
        for file in files:
            path = os.path.join(root, file)
            image = cv2.imread(path)
            gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
            image_array = np.array(gray, "uint8")
            label = os.path.basename(root)

            if label not in label_ids:
                label_ids = current_id
                current_id += 1
            id_ = label_ids

            faces = face_cascade.detectMultiScale(image_array, scaleFactor=1.5, minNeighbors=5)

            for (x, y, w, h) in faces:
                roi = image_array[y:y + h, x:x + w]
                x_train.append(roi)
                y_labels.append(id_)
    with open("label.pickle", "wb") as f:
        pickle.dump(label_ids, f)
        print(label_ids)
    recognizer.train(x_train, np.array(y_labels))
    recognizer.save("my_trainer.xml")

# while True:
name = input("please input name :")
if len(name) != 0:
    input_face(name)
train_face()
face_recognition()

cap.release()
cv2.destroyAllWindows()

4、树莓派上使用iic