from deepface import DeepFace

# 分析图像中的情绪
results = DeepFace.analyze(img_path="person.jpg", actions=['emotion'])
print(results)

# 输出示例：
# [{'region': {'x': 170, 'y': 105, 'w': 200, 'h': 200}, 
#   'emotion': {'angry': 0.03, 'disgust': 0.01, 'fear': 0.02, 'happy': 85.21, 'sad': 0.05, 'surprise': 0.03, 'neutral': 14.65}}]

# 同时分析多种属性
results = DeepFace.analyze(img_path="person.jpg", 
                         actions=['emotion', 'age', 'gender', 'race'])
print(results)

import cv2
from deepface import DeepFace

def real_time_emotion_detection():
    # 初始化摄像头
    cap = cv2.VideoCapture(0)
    
    print("实时情绪检测系统已启动，按 'q' 键退出")
    
    while True:
        # 读取摄像头帧
        ret, frame = cap.read()
        
        if not ret:
            break
            
        try:
            # 检测情绪
            results = DeepFace.analyze(frame, actions=['emotion'], detector_backend='opencv')
            
            # 处理单个检测结果
            if isinstance(results, dict):
                results = [results]
                
            # 在图像上绘制检测结果
            for result in results:
                # 获取边界框坐标
                region = result["region"]
                x, y, w, h = region["x"], region["y"], region["w"], region["h"]
                
                # 获取情绪数据
                emotions = result["emotion"]
                
                # 找到置信度最高的情绪
                dominant_emotion = max(emotions, key=emotions.get)
                confidence = emotions[dominant_emotion]
                
                # 绘制边界框
                cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
                
                # 在边界框上方显示情绪和置信度
                text = f"{dominant_emotion}: {confidence:.2f}"
                cv2.putText(frame, text, (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 
                           0.9, (0, 255, 0), 2)
        except Exception as e:
            # 如果检测不到人脸，继续下一帧
            pass
        
        # 显示结果
        cv2.imshow('实时情绪检测', frame)
        
        # 按 'q' 键退出
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    
    # 释放资源
    cap.release()
    cv2.destroyAllWindows()

# 运行实时情绪检测系统
if __name__ == "__main__":
    real_time_emotion_detection()

import cv2
from deepface import DeepFace
from collections import deque

class EnhancedEmotionDetector:
    def __init__(self, max_history=100):
        self.cap = cv2.VideoCapture(0)
        self.emotion_history = deque(maxlen=max_history)
        self.emotions_list = ['angry', 'disgust', 'fear', 'happy', 'sad', 'surprise', 'neutral']
        
    def run(self):
        print("增强版情绪检测系统已启动，按 'q' 键退出")
        
        while True:
            ret, frame = self.cap.read()
            
            if not ret:
                break
                
            try:
                # 检测情绪
                results = DeepFace.analyze(frame, actions=['emotion'], detector_backend='opencv')
                
                # 处理单个检测结果
                if isinstance(results, dict):
                    results = [results]
                
                # 处理检测结果
                for result in results:
                    region = result["region"]
                    x, y, w, h = region["x"], region["y"], region["w"], region["h"]
                    emotions = result["emotion"]
                    
                    # 找到主导情绪
                    dominant_emotion = max(emotions, key=emotions.get)
                    confidence = emotions[dominant_emotion]
                    
                    # 添加到历史记录
                    self.emotion_history.append(dominant_emotion)
                    
                    # 绘制边界框和情绪信息
                    cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
                    text = f"{dominant_emotion}: {confidence:.2f}"
                    cv2.putText(frame, text, (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 
                               0.9, (0, 255, 0), 2)
                
                # 显示统计信息
                if len(self.emotion_history) > 0:
                    self.display_statistics(frame)
            except Exception as e:
                # 如果检测不到人脸，继续下一帧
                pass
            
            cv2.imshow('增强版情绪检测系统', frame)
            
            if cv2.waitKey(1) & 0xFF == ord('q'):
                break
                
        self.cleanup()
        
    def display_statistics(self, frame):
        if len(self.emotion_history) == 0:
            return
            
        # 计算情绪分布
        emotion_counts = {emotion: 0 for emotion in self.emotions_list}
        for emotion in self.emotion_history:
            emotion_counts[emotion] += 1
            
        total = len(self.emotion_history)
        dominant_overall = max(emotion_counts, key=emotion_counts.get)
        
        # 在图像上显示统计信息
        cv2.putText(frame, f"主导情绪：{dominant_overall}", (10, 30), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
        cv2.putText(frame, f"检测次数：{total}", (10, 60), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
        
    def cleanup(self):
        self.cap.release()
        cv2.destroyAllWindows()

# 运行增强版情绪检测系统
if __name__ == "__main__":
    detector = EnhancedEmotionDetector()
    detector.run()

from deepface import DeepFace

# 对视频流进行实时分析
DeepFace.stream(db_path="./database", actions=['emotion'])

from fer import FER
import cv2

# 读取图像
img = cv2.imread("person.jpg")

# 创建检测器实例
detector = FER()

# 检测情绪
results = detector.detect_emotions(img)
print(results)

# 输出示例：
# [{'box': [277, 90, 48, 63], 'emotions': {'angry': 0.02, 'disgust': 0.0, 'fear': 0.05, 'happy': 0.16, 'neutral': 0.09, 'sad': 0.27, 'surprise': 0.41}}]

# 获取最主要的情绪
emotion, score = detector.top_emotion(img)
print(f"主要情绪：{emotion}, 置信度：{score}")

import cv2
from fer import FER

def real_time_emotion_detection():
    # 初始化摄像头
    cap = cv2.VideoCapture(0)
    
    # 创建情绪检测器，使用 MTCNN 提高检测精度
    detector = FER(mtcnn=True)
    
    print("实时情绪检测系统已启动，按 'q' 键退出")
    
    while True:
        # 读取摄像头帧
        ret, frame = cap.read()
        
        if not ret:
            break
            
        # 检测情绪
        results = detector.detect_emotions(frame)
        
        # 在图像上绘制检测结果
        for result in results:
            # 获取边界框坐标
            x, y, w, h = result["box"]
            
            # 获取情绪数据
            emotions = result["emotions"]
            
            # 找到置信度最高的情绪
            dominant_emotion = max(emotions, key=emotions.get)
            confidence = emotions[dominant_emotion]
            
            # 绘制边界框
            cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
            
            # 在边界框上方显示情绪和置信度
            text = f"{dominant_emotion}: {confidence:.2f}"
            cv2.putText(frame, text, (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 
                       0.9, (0, 255, 0), 2)
        
        # 显示结果
        cv2.imshow('实时情绪检测', frame)
        
        # 按 'q' 键退出
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    
    # 释放资源
    cap.release()
    cv2.destroyAllWindows()

# 运行实时情绪检测系统
if __name__ == "__main__":
    real_time_emotion_detection()

import cv2
import numpy as np
from fer import FER
from collections import deque
import matplotlib.pyplot as plt

class EnhancedEmotionDetector:
    def __init__(self, max_history=100):
        self.cap = cv2.VideoCapture(0)
        self.detector = FER(mtcnn=True)
        self.emotion_history = deque(maxlen=max_history)
        self.emotions_list = ['angry', 'disgust', 'fear', 'happy', 'neutral', 'sad', 'surprise']
        
    def run(self):
        print("增强版情绪检测系统已启动，按 'q' 键退出")
        
        while True:
            ret, frame = self.cap.read()
            
            if not ret:
                break
                
            # 检测情绪
            results = self.detector.detect_emotions(frame)
            
            # 处理检测结果
            for result in results:
                x, y, w, h = result["box"]
                emotions = result["emotions"]
                
                # 找到主导情绪
                dominant_emotion = max(emotions, key=emotions.get)
                confidence = emotions[dominant_emotion]
                
                # 添加到历史记录
                self.emotion_history.append(dominant_emotion)
                
                # 绘制边界框和情绪信息
                cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
                text = f"{dominant_emotion}: {confidence:.2f}"
                cv2.putText(frame, text, (x, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 
                           0.9, (0, 255, 0), 2)
            
            # 显示统计信息
            if len(self.emotion_history) > 0:
                self.display_statistics(frame)
            
            cv2.imshow('增强版情绪检测系统', frame)
            
            if cv2.waitKey(1) & 0xFF == ord('q'):
                break
                
        self.cleanup()
        
    def display_statistics(self, frame):
        if len(self.emotion_history) == 0:
            return
            
        # 计算情绪分布
        emotion_counts = {emotion: 0 for emotion in self.emotions_list}
        for emotion in self.emotion_history:
            emotion_counts[emotion] += 1
            
        total = len(self.emotion_history)
        dominant_overall = max(emotion_counts, key=emotion_counts.get)
        
        # 在图像上显示统计信息
        cv2.putText(frame, f"主导情绪：{dominant_overall}", (10, 30), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
        cv2.putText(frame, f"检测次数：{total}", (10, 60), 
                   cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 255, 255), 2)
        
    def cleanup(self):
        self.cap.release()
        cv2.destroyAllWindows()

# 运行增强版情绪检测系统
if __name__ == "__main__":
    detector = EnhancedEmotionDetector()
    detector.run()

from fer import Video
from fer import FER

def analyze_video_emotions(video_filename):
    # 创建视频分析器
    video = Video(video_filename)
    
    # 创建情绪检测器
    detector = FER(mtcnn=True)
    
    # 分析视频
    raw_data = video.analyze(detector, display=True)
    
    # 转换为 pandas DataFrame 便于分析
    df = video.to_pandas(raw_data)
    
    # 保存结果到 CSV 文件
    df.to_csv("emotion_analysis.csv", index=False)
    
    return df

# 使用示例
# df = analyze_video_emotions("sample_video.mp4")