1 Star 32 Fork 16

PPOV_NUC / yolov5_infer

加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
克隆/下载
openvino2022-device-for-mqtt.py 7.63 KB
一键复制 编辑 原始数据 按行查看 历史
#version: python 3.x 测试过
#用于模拟一个基于mqtt协议传输OpenVINO推理结果
import paho.mqtt.client as mqtt
import json
import time
import queue as Queue
import threading,random
import cv2
import numpy as np
import time
import yaml
# from openvino.inference_engine import IECore # the version of openvino <= 2021.4.2
from openvino.runtime import Core # the version of openvino >= 2022.1
BROKER_HOST_ADDR = "localhost"
BROKER_HOST_PORT = 1883
USERNAME = "huaqiaoz" #非安全模式忽略用户名秘密
PWD = "1234" #非安全模式忽略用户名秘密
#cmd topic本质上就是你的设备监听的topic,
#也是在UI上添加device的时候,地址中所填数据,和用户名密码等一起组成当前设备的唯一标识。
CMD_TOPIC = "CommandTopic"
RESPONSE_TOPIC = "ResponseTopic"
DATA_TOPIC = "DataTopic"
# 载入COCO Label
with open('./coco.yaml','r', encoding='utf-8') as f:
result = yaml.load(f.read(),Loader=yaml.FullLoader)
class_list = result['names']
# YOLOv5s输入尺寸
INPUT_WIDTH = 640
INPUT_HEIGHT = 640
# 载入yolov5s onnx模型
model_path = "./yolov5s.onnx"
# Read yolov5s onnx model with OpenVINO API
ie = Core() #Initialize IECore
net = ie.compile_model(model=model_path, device_name="AUTO")
# 开启Webcam,并设置为1280x720
cap = cv2.VideoCapture(0)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)
# 调色板
colors = [(255, 255, 0), (0, 255, 0), (0, 255, 255), (255, 0, 0)]
# 基于OpenVINO的YOLOv5s.onnx推理计算子函数
# 目标检测函数,返回检测结果
def detect(image, net):
blob = cv2.dnn.blobFromImage(image, 1 / 255.0, (INPUT_WIDTH, INPUT_HEIGHT), swapRB=True, crop=False)
#result = net.infer({"images": blob})
#preds = result["output"]
preds = net([blob])[next(iter(net.outputs))] # API version>=2022.1
return preds
# YOLOv5的后处理函数,解析模型的输出
def wrap_detection(input_image, output_data):
class_ids = []
confidences = []
boxes = []
#print(output_data.shape)
rows = output_data.shape[0]
image_width, image_height, _ = input_image.shape
x_factor = image_width / INPUT_WIDTH
y_factor = image_height / INPUT_HEIGHT
for r in range(rows):
row = output_data[r]
confidence = row[4]
if confidence >= 0.4:
classes_scores = row[5:]
_, _, _, max_indx = cv2.minMaxLoc(classes_scores)
class_id = max_indx[1]
if (classes_scores[class_id] > .25):
confidences.append(confidence)
class_ids.append(class_id)
x, y, w, h = row[0].item(), row[1].item(), row[2].item(), row[3].item()
left = int((x - 0.5 * w) * x_factor)
top = int((y - 0.5 * h) * y_factor)
width = int(w * x_factor)
height = int(h * y_factor)
box = np.array([left, top, width, height])
boxes.append(box)
indexes = cv2.dnn.NMSBoxes(boxes, confidences, 0.25, 0.45)
result_class_ids = []
result_confidences = []
result_boxes = []
for i in indexes:
result_confidences.append(confidences[i])
result_class_ids.append(class_ids[i])
result_boxes.append(boxes[i])
return result_class_ids, result_confidences, result_boxes
# 按照YOLOv5要求,先将图像长:宽 = 1:1,多余部分填充黑边
def format_yolov5(frame):
row, col, _ = frame.shape
_max = max(col, row)
result = np.zeros((_max, _max, 3), np.uint8)
result[0:row, 0:col] = frame
return result
globalQueue = Queue.Queue()
def gen():
return round(random.uniform(0, 50),2)
def send_data():
#java版本, name的值为添加的设备名
#data = {"randnum":520.1314,"name":"mqtt-device-01"}
#go版本, name的值为添加的设备名(required), go版本的区别是必须带上cmd字段
# data = {"randnum":520.1314,"name":"mqtt-device-01","cmd":"randnum","method":"get"}
#v2.1.0版本数据格式
#数据格式 data = {"randnum":520.1314,"name":"mqtt-device-01","cmd":"randnum","method":"get"}
data = {"randnum":gen(),"name":"mqtt-device-01","cmd":"randnum", "method":"get"}
print("sending data actively! " + json.dumps(data))
client.publish(DATA_TOPIC,json.dumps(data) , qos=0, retain=False)
class SendDataActiveServer(threading.Thread):
def __init__(self,threadID,name,queue):
super(SendDataActiveServer,self).__init__()
self.threadID = threadID
self.name = name
self.queue = queue
self.active = False
def run(self):
while 1==1 :
#v2.1.0 接收到的bool值自动自动转换为python能识别的bool值,无须二次转换
if self.active:
send_data()
time.sleep(1)
self.getItemFromQueue()
else:
time.sleep(1)
self.getItemFromQueue()
def getItemFromQueue(self):
try:
if self.queue.get(block=False):
self.active = True
else:
self.active = False
except Queue.Empty:
#quene.get()方法在队列中为空是返回异常,捕获异常什么都不做,保持active原状
time.sleep(0.1)
#当接收到命令,响应命令
#v2.1.0 get命令格式: {"cmd":"ping","method":"get","uuid":"f46ae5c7-2a08-4f56-a38b-d9e122d255d0"}
#v2.1.0 set命令格式 {"cmd":"ping","ping":"ping_set_value","method":"set","uuid":"95d559ad-7140-4d62-8dd8-6465c37fec10"}
def on_message(client, userdata, msg):
print(msg.topic+" "+str(msg.payload)+'\n')
d = json.loads(msg.payload)
if d['cmd'] == "message":
print("This is message cmd")
if d['method'] == "get": # 执行OpenVINO推理程序,并拿到结果
_, frame = cap.read()
if frame is None:
print("End of stream")
# 将图像按最大边1:1放缩
inputImage = format_yolov5(frame)
# 执行推理计算
outs = detect(inputImage, net)
# 拆解推理结果
class_ids, confidences, boxes = wrap_detection(inputImage, outs[0])
msg = ""
for (classid, confidence, box) in zip(class_ids, confidences, boxes):
msg += f"detected: {class_list[classid]}:{confidence:.2f}, at xmin:{box[0]}, ymin:{box[1]}, xmax:{box[2]}, ymax:{box[3]}.\n"
d['message'] = msg
print(msg)
elif d['method'] == "set":
d['result'] = "set successed."
if d['cmd'] == "ping":
print("This is ping cmd")
d['ping'] = "pong"
if d['cmd'] == "randnum":
print("This is randnum cmd")
d['randnum'] = gen()
if d['cmd'] == "collect" and d['method'] == "set":
print("This is collect set cmd")
d['result'] = "set successed."
#param的值是true或false,且是字符串类型,
#globalQueue.put(d['param'])
#v2.1.0命令格式中移除了param属性,使用device resource name为实际参数名
globalQueue.put(d['collect'])
elif d['cmd'] == "collect" and d['method'] == "get":
print("This is collect get cmd")
d['collect'] = thread.active
print(json.dumps(d))
client.publish(RESPONSE_TOPIC, json.dumps(d))
def on_connect(client, userdata, flags, rc):
print("Connected success with result code "+str(rc))
#监听命令
client.subscribe(CMD_TOPIC)
client = mqtt.Client()
client.username_pw_set(USERNAME, PWD)
client.on_message = on_message
client.on_connect = on_connect
client.connect(BROKER_HOST_ADDR, BROKER_HOST_PORT, 60)
#开始独立线程用于主动发送数据
thread = SendDataActiveServer("Thread-1", "SendDataServerThread", globalQueue)
thread.setDaemon(True)
thread.start()
client.loop_forever()
Python
1
https://gitee.com/ppov-nuc/yolov5_infer.git
git@gitee.com:ppov-nuc/yolov5_infer.git
ppov-nuc
yolov5_infer
yolov5_infer
main

搜索帮助