如何在深度学习目标检测中建立基于YOLOv8和DeepSort的智能寻人人员检测系统，行人重识别（ReID）和目标跟踪，QT界面进行展示

基于YOLOv5+Deepsort的智能寻人系统 行人重识别+目标跟踪 QT界面

以下是一个完整的流程，从安装依赖、准备数据集、配置YOLOv8、训练模型、评估模型、推理代码到构建GUI应用程序。
仅供参考，

1. 安装依赖

确保已经安装了必要的Python库：

pip install ultralytics opencv-python-headless numpy torch torchvision scikit-learn

2. 准备数据集

数据集转换：VOC格式 → YOLO格式

假设你已经有了标注好的数据集，以下是转换脚本：

import os
import xml.etree.ElementTree as ET

def convert(size, box):
    dw = 1. / size[0]
    dh = 1. / size[1]
    x = (box[0] + box[1]) / 2.0 - 1
    y = (box[2] + box[3]) / 2.0 - 1
    w = box[1] - box[0]
    h = box[3] - box[2]
    x = x * dw
    w = w * dw
    y = y * dh
    h = h * dh
    return (x, y, w, h)

def convert_annotation(xml_file, output_file, classes):
    tree = ET.parse(xml_file)
    root = tree.getroot()
    size = root.find('size')
    w = int(size.find('width').text)
    h = int(size.find('height').text)

    with open(output_file, 'w') as out_file:
        for obj in root.iter('object'):
            cls = obj.find('name').text
            if cls not in classes:
                continue
            cls_id = classes.index(cls)
            xmlbox = obj.find('bndbox')
            b = (float(xmlbox.find('xmin').text), float(xmlbox.find('xmax').text),
                 float(xmlbox.find('ymin').text), float(xmlbox.find('ymax').text))
            bb = convert((w, h), b)
            out_file.write(f"{cls_id} {' '.join(map(str, bb))}\n")

# 配置路径和类别
classes = ['person']
xml_dir = './path/to/voc/annotations'
output_dir = './path/to/yolo/labels'

os.makedirs(output_dir, exist_ok=True)

for xml_file in os.listdir(xml_dir):
    if xml_file.endswith('.xml'):
        convert_annotation(os.path.join(xml_dir, xml_file),
                           os.path.join(output_dir, xml_file.replace('.xml', '.txt')),
                           classes)

创建data.yaml配置文件

创建一个data.yaml文件，描述数据集路径和类别信息：

train: ./path/to/train/images
val: ./path/to/val/images
test: ./path/to/test/images

nc: 1
names: ['person']

3. 配置YOLOv8并训练模型

加载YOLOv8模型并开始训练

from ultralytics import YOLO

# 加载YOLOv8模型
model = YOLO('yolov8n.yaml')  # 或者使用预训练权重：'yolov8n.pt'

# 开始训练
results = model.train(data='./path/to/data.yaml',
                      epochs=100,
                      imgsz=640,  # 输入图像大小
                      batch=16)   # 批次大小

模型评估

训练完成后，可以对验证集进行评估：

metrics = model.val()  # 使用验证集评估模型
print(f"Validation mAP: {metrics.box.map}")

4. 推理代码

编写推理代码，用于检测单张图片中的行人：

from PIL import Image
import cv2
import torch
from deep_sort.deep_sort import DeepSort

def detect_image(image_path, model, deepsort):
    results = model.predict(source=image_path, save=False)  # 进行推理
    image = cv2.imread(image_path)
    detections = []
    for pred in results:
        boxes = pred.boxes.xyxy.cpu().numpy()
        confs = pred.boxes.conf.cpu().numpy()
        cls_ids = pred.boxes.cls.cpu().numpy()
        for box, conf, cls_id in zip(boxes, confs, cls_ids):
            if conf > 0.5:  # 设置置信度阈值
                detections.append((*box, conf))

    # 使用DeepSort进行跟踪
    outputs = deepsort.update(detections, image)

    for track in outputs:
        x1, y1, x2, y2, id = track
        label = f'ID: {int(id)}'
        cv2.rectangle(image, (x1, y1), (x2, y2), (0, 255, 0), 2)
        cv2.putText(image, label, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)
    
    return image

# 示例用法
image_path = './path/to/test/image.jpg'
model = YOLO('./path/to/best.pt')  # 加载训练好的模型
deepsort = DeepSort("deep_sort/deep/checkpoint/ckpt.t7")  # 初始化DeepSort

detected_image = detect_image(image_path, model, deepsort)
cv2.imshow('Detected Image', detected_image)
cv2.waitKey(0)
cv2.destroyAllWindows()

5. 构建GUI应用程序

使用PyQt5构建一个简单的GUI应用程序，用于选择图片并展示检测结果：

import sys
from PyQt5.QtWidgets import QApplication, QWidget, QVBoxLayout, QPushButton, QLabel, QFileDialog
from PyQt5.QtGui import QPixmap
from PyQt5.QtCore import Qt
import cv2
from ultralytics import YOLO
from deep_sort.deep_sort import DeepSort

class App(QWidget):
    def __init__(self):
        super().__init__()
        self.initUI()

        # 加载模型
        self.model = YOLO('./path/to/best.pt')
        self.deepsort = DeepSort("deep_sort/deep/checkpoint/ckpt.t7")

    def initUI(self):
        self.setWindowTitle('智能寻人系统')

        layout = QVBoxLayout()

        self.image_label = QLabel(self)
        layout.addWidget(self.image_label)

        self.select_button = QPushButton('选择图片', self)
        self.select_button.clicked.connect(self.select_image)
        layout.addWidget(self.select_button)

        self.setLayout(layout)
        self.show()

    def select_image(self):
        options = QFileDialog.Options()
        file_name, _ = QFileDialog.getOpenFileName(self, "选择图片", "", "Image Files (*.jpg *.jpeg *.png)", options=options)
        if file_name:
            self.detect_and_show(file_name)

    def detect_and_show(self, image_path):
        image = cv2.imread(image_path)
        detections = []
        results = self.model.predict(source=image_path, save=False)
        for pred in results:
            boxes = pred.boxes.xyxy.cpu().numpy()
            confs = pred.boxes.conf.cpu().numpy()
            cls_ids = pred.boxes.cls.cpu().numpy()
            for box, conf, cls_id in zip(boxes, confs, cls_ids):
                if conf > 0.5:
                    detections.append((*box, conf))

        outputs = self.deepsort.update(detections, image)

        for track in outputs:
            x1, y1, x2, y2, id = track
            label = f'ID: {int(id)}'
            cv2.rectangle(image, (x1, y1), (x2, y2), (0, 255, 0), 2)
            cv2.putText(image, label, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)

        height, width, channel = image.shape
        bytes_per_line = 3 * width
        q_img = QImage(image.data, width, height, bytes_per_line, QImage.Format_RGB888).rgbSwapped()
        pixmap = QPixmap.fromImage(q_img)
        self.image_label.setPixmap(pixmap)

if __name__ == '__main__':
    app = QApplication(sys.argv)
    ex = App()
    sys.exit(app.exec_())

6. 运行

1. 确保所有路径和参数正确无误。
2. 运行上述代码，启动GUI应用程序。
3. 在GUI中选择一张测试图片，程序会自动检测行人并显示结果。

总结 仅供参考

安装依赖、准备数据集、配置YOLOv8、训练模型、评估模型、推理代码到构建GUI应用程序的完整流程。同学，长点心，请根据实际需求调整路径、参数和代码细节。

仅供参考。