输出结果

设计思路

核心代码

from statistics import mode

import cv2

from keras.models import load_model

import numpy as np

detection_model_path = '../trained_models/detection_models/haarcascade_frontalface_default.xml'

emotion_model_path = '../trained_models/emotion_models/fer2013_mini_XCEPTION.102-0.66.hdf5'

gender_model_path = '../trained_models/gender_models/simple_CNN.81-0.96.hdf5'

emotion_labels = get_labels('fer2013')

gender_labels = get_labels('imdb')

font = cv2.FONT_HERSHEY_SIMPLEX

frame_window = 10  

gender_offsets = (30, 60)

emotion_offsets = (20, 40)

face_detection = load_detection_model(detection_model_path)

emotion_classifier = load_model(emotion_model_path, compile=False)

gender_classifier = load_model(gender_model_path, compile=False)

emotion_target_size = emotion_classifier.input_shape[1:3]

gender_target_size = gender_classifier.input_shape[1:3]

gender_window = []

emotion_window = []

cv2.namedWindow('window_frame_by_Jason_Niu')

# video_capture = cv2.VideoCapture(0)

video_capture = cv2.VideoCapture("F:\File_Python\Python_example\YOLOv3_use_TF\RunMan5.mp4")

while True:

   bgr_image = video_capture.read()[1]

   gray_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2GRAY) #分别将读取的图像进行灰化、RGB化处理

   rgb_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2RGB)

   faces = detect_faces(face_detection, gray_image)  

   for face_coordinates in faces:

   

       x1, x2, y1, y2 = apply_offsets(face_coordinates, gender_offsets)

       rgb_face = rgb_image[y1:y2, x1:x2]  

       x1, x2, y1, y2 = apply_offsets(face_coordinates, emotion_offsets)

       gray_face = gray_image[y1:y2, x1:x2]

       try:

           rgb_face = cv2.resize(rgb_face, (gender_target_size))

           gray_face = cv2.resize(gray_face, (emotion_target_size))

       except:

           continue

       gray_face = preprocess_input(gray_face, False)  

       gray_face = np.expand_dims(gray_face, 0)  

       gray_face = np.expand_dims(gray_face, -1)

       emotion_label_arg = np.argmax(emotion_classifier.predict(gray_face))

       emotion_text = emotion_labels[emotion_label_arg]

       emotion_window.append(emotion_text)    

       rgb_face = np.expand_dims(rgb_face, 0)

       rgb_face = preprocess_input(rgb_face, False)

       gender_prediction = gender_classifier.predict(rgb_face)

       gender_label_arg = np.argmax(gender_prediction)

       gender_text = gender_labels[gender_label_arg]

       gender_window.append(gender_text)

       if len(gender_window) > frame_window:

           emotion_window.pop(0)        

           gender_window.pop(0)

       try:

           emotion_mode = mode(emotion_window)

           gender_mode = mode(gender_window)

       except:

           continue

       if gender_text == gender_labels[0]:

           color = (0, 0, 255)

       else:

           color = (255, 0, 0)

       draw_bounding_box(face_coordinates, rgb_image, color)

       draw_text(face_coordinates, rgb_image, gender_mode,  

                 color, 0, -20, 1, 4)

       draw_text(face_coordinates, rgb_image, emotion_mode,

                 color, 0, -45, 1, 4)

       

   bgr_image = cv2.cvtColor(rgb_image, cv2.COLOR_RGB2BGR)  

   cv2.namedWindow("window_frame_by_Jason_Niu",0);

   cv2.resizeWindow("window_frame_by_Jason_Niu", 640, 380);

   cv2.imshow('window_frame_by_Jason_Niu', bgr_image)

   if cv2.waitKey(1) & 0xFF == ord('q'):

       break