基于R-CNN深度学习的行人车辆检测与计数系统：数据集、模型和UI界面的完整实现

 dataset/

    
 ├── images/
    
 │   ├── image1.jpg
    
 │   ├── image2.jpg
    
 │   └── ...
    
 └── annotations/
    
     ├── image1.xml
    
     ├── image2.xml
    
     └── ...
    
    
    
    
    代码解读

 import os

    
 import numpy as np
    
 import cv2
    
 import xml.etree.ElementTree as ET
    
  
    
 def load_data(images_dir, annotations_dir):
    
     images = []
    
     labels = []
    
     
    
     for filename in os.listdir(images_dir):
    
     if filename.endswith('.jpg'):
    
         image_path = os.path.join(images_dir, filename)
    
         images.append(cv2.imread(image_path))
    
         annotation_path = os.path.join(annotations_dir, filename.replace('.jpg', '.xml'))
    
         labels.append(parse_annotation(annotation_path))
    
     
    
     return np.array(images), labels
    
  
    
 def parse_annotation(annotation_path):
    
     tree = ET.parse(annotation_path)
    
     root = tree.getroot()
    
     boxes = []
    
     for obj in root.findall('object'):
    
     class_name = obj.find('name').text
    
     bbox = obj.find('bndbox')
    
     xmin = int(bbox.find('xmin').text)
    
     ymin = int(bbox.find('ymin').text)
    
     xmax = int(bbox.find('xmax').text)
    
     ymax = int(bbox.find('ymax').text)
    
     boxes.append((class_name, xmin, ymin, xmax, ymax))
    
     return boxes
    
    
    
    
    代码解读

 from keras.preprocessing.image import ImageDataGenerator

    
  
    
 def augment_data(images):
    
     datagen = ImageDataGenerator(
    
     rotation_range=20,
    
     width_shift_range=0.2,
    
     height_shift_range=0.2,
    
     shear_range=0.2,
    
     zoom_range=0.2,
    
     horizontal_flip=True,
    
     fill_mode='nearest'
    
     )
    
     
    
     augmented_images = []
    
     for img in images:
    
     img = img.reshape((1,) + img.shape)
    
     for batch in datagen.flow(img, batch_size=1):
    
         augmented_images.append(batch[0].astype(np.uint8))
    
         break
    
     
    
     return np.array(augmented_images)
    
    
    
    
    代码解读

 import tensorflow as tf

    
 from tensorflow.keras import layers, models
    
  
    
 def create_rcnn_model(num_classes):
    
     base_model = tf.keras.applications.VGG16(weights='imagenet', include_top=False, input_shape=(224, 224, 3))
    
     base_model.trainable = False  # 冻结卷积基
    
     x = base_model.output
    
     x = layers.Flatten()(x)
    
     x = layers.Dense(512, activation='relu')(x)
    
     x = layers.Dropout(0.5)(x)
    
     predictions = layers.Dense(num_classes, activation='softmax')(x)
    
  
    
     model = models.Model(inputs=base_model.input, outputs=predictions)
    
     model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
    
     return model
    
    
    
    
    代码解读

 from sklearn.model_selection import train_test_split

    
 from sklearn.preprocessing import LabelEncoder
    
  
    
 # 数据加载
    
 images_dir = 'path/to/dataset/images'  # 替换为数据集路径
    
 annotations_dir = 'path/to/dataset/annotations'  # 替换为数据集路径
    
 X, y = load_data(images_dir, annotations_dir)
    
  
    
 # 标签编码
    
 label_encoder = LabelEncoder()
    
 y_encoded = [label_encoder.fit_transform([label[0] for label in labels]) for labels in y]
    
  
    
 # 数据集分割
    
 X_train, X_val, y_train, y_val = train_test_split(X, y_encoded, test_size=0.2, random_state=42)
    
  
    
 # 创建模型
    
 num_classes = len(label_encoder.classes_)
    
 model = create_rcnn_model(num_classes)
    
  
    
 # 训练模型
    
 history = model.fit(X_train, y_train, validation_data=(X_val, y_val), epochs=20, batch_size=32)
    
    
    
    
    代码解读

 import tkinter as tk

    
 from tkinter import filedialog
    
 from PIL import Image, ImageTk
    
  
    
 class ObjectDetectionApp:
    
     def __init__(self, master):
    
     self.master = master
    
     master.title("行人车辆检测与计数系统")
    
     
    
     self.label = tk.Label(master, text="请上传视频进行检测")
    
     self.label.pack()
    
     
    
     self.upload_button = tk.Button(master, text="上传视频", command=self.upload_video)
    
     self.upload_button.pack()
    
     
    
     self.detect_button = tk.Button(master, text="检测", command=self.detect_objects)
    
     self.detect_button.pack()
    
     
    
     self.result_label = tk.Label(master, text="")
    
     self.result_label.pack()
    
     
    
     self.video_label = tk.Label(master)
    
     self.video_label.pack()
    
  
    
     def upload_video(self):
    
     self.video_path = filedialog.askopenfilename()
    
     self.result_label.config(text="视频已上传，可以进行检测！")
    
  
    
     def detect_objects(self):
    
     cap = cv2.VideoCapture(self.video_path)
    
     while cap.isOpened():
    
         ret, frame = cap.read()
    
         if not ret:
    
             break
    
         # 在这里添加检测逻辑
    
         # detection_results = model.predict(frame)
    
         # 处理检测结果并绘制到图像上
    
  
    
         # 显示检测结果
    
         img = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    
         img = Image.fromarray(img)
    
         img = ImageTk.PhotoImage(img)
    
         self.video_label.img = img
    
         self.video_label.config(image=img)
    
  
    
     cap.release()
    
     self.result_label.config(text="检测完成！")
    
  
    
 if __name__ == "__main__":
    
     root = tk.Tk()
    
     app = ObjectDetectionApp(root)
    
     root.mainloop()
    
    
    
    
    代码解读

 from sklearn.metrics import classification_report

    
  
    
 y_pred = model.predict(X_val)
    
 y_pred_classes = np.argmax(y_pred, axis=1)
    
  
    
 print(classification_report(y_val, y_pred_classes, target_names=label_encoder.classes_))
    
    
    
    
    代码解读

 def draw_boxes(frame, boxes):

    
     for box in boxes:
    
     class_name, xmin, ymin, xmax, ymax = box
    
     cv2.rectangle(frame, (xmin, ymin), (xmax, ymax), (0, 255, 0), 2)
    
     cv2.putText(frame, class_name, (xmin, ymin - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
    
     return frame
    
    
    
    
    代码解读