opencv—人脸识别、分类

备注：
‘’’
一.training-data 存放训练照片的文件夹
training-data/s1 放第一个人的照片各种表情状态是否佩戴眼镜等情况的照片
training-data/s2是放第二人的照片各种表情状态是否佩戴眼镜等情况的照片
以此类推
有多少人建多少个文件夹
二、test-data文件夹放置的是要识别人脸的照片
三、face_cascade ：lbpcascade_frontalface.xml放置的路径
‘’’

代码1：

# coding: utf-8
import cv2
import os
import numpy as np
import shutil
import glob

subjects = ["a", "b","c","d","e"]    #预测

def detect_face(img):
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    face_cascade = cv2.CascadeClassifier('../cv-class/lbpcascade_frontalface.xml')   #opencv自带的分类器
    faces = face_cascade.detectMultiScale(gray, scaleFactor=1.2, minNeighbors=1)
    if (len(faces) == 0):
        return None, None
    (x, y, w, h) = faces[0]
    return gray[y:y + w, x:x + h], faces[0]

def prepare_training_data(data_folder_path):
    dirs = os.listdir(data_folder_path)
    faces = []
    labels = []
    for dir_name in dirs:
        if not dir_name.startswith("s"):
            continue;
        label = int(dir_name.replace("s", ""))   

        subject_dir_path = data_folder_path + "/" + dir_name
        subject_images_names = os.listdir(subject_dir_path)

        for image_name in subject_images_names:
            if image_name.startswith("."):
                continue;
            image_path = subject_dir_path + "/" + image_name
            image = cv2.imread(image_path)
            cv2.imshow("Training on image...", cv2.resize(image, (400, 500)))
            cv2.waitKey(100)

            face, rect = detect_face(image)
            if face is not None:
                faces.append(face)
                labels.append(label)
    cv2.waitKey(1)
    cv2.destroyAllWindows()
    return faces, labels

print("Preparing data...")
faces, labels = prepare_training_data("training-data")  #调用训练库
print("Total faces: ", len(faces))
print("Total labels: ", len(labels))

face_recognizer = cv2.face.createLBPHFaceRecognizer()
face_recognizer.train(faces, np.array(labels))          #开始训练模型

def draw_rectangle(img, rect):
    (x, y, w, h) = rect
    cv2.rectangle(img, (x, y), (x + w, y + h), (0, 255, 0), 2)

def draw_text(img, text, x, y):
    cv2.putText(img, text, (x, y), cv2.FONT_HERSHEY_PLAIN, 1.5, (0, 255, 0), 2)

def predict(test_img):
    img = test_img.copy()
    face, rect = detect_face(img)
    label = face_recognizer.predict(face)
    label_text = subjects[label-1]

    if label_text == 'a':
        print(label_text, '1')
        shutil.copy(file,'img2/1/'+file.split('\\')[-1])
    if label_text == 'b':
        print(label_text, '2')
        shutil.copy(file,'img2/2/'+file.split('\\')[-1])
    if label_text == 'c':
        print(label_text, '3')
        shutil.copy(file, 'img2/3/' + file.split('\\')[-1])
    if label_text == 'd':
        print(label_text, '4')
        shutil.copy(file, 'img2/4/' + file.split('\\')[-1])
    if label_text == 'e':
        print(label_text, '5')
        shutil.copy(file, 'img2/5/' + file.split('\\')[-1])

    draw_rectangle(img, rect)
    draw_text(img, label_text, rect[0], rect[1] - 5)
    return img


files  = glob.glob('img/*.jpg')
for file in files:
    print(file)
    test_img = cv2.imread(file)
    predicted_img = predict(test_img)

    # cv2.imshow(subjects[1], cv2.resize(predicted_img, (400, 500)))
    # cv2.waitKey(1)
    # cv2.destroyAllWindows()

得到结果如下，分类正确：


省略剩下三个文件夹。

---

---

代码2—自己训练模型，进行识别人类：

# coding: utf-8
import cv2
import os
import numpy as np
import shutil
import glob

subjects = ["a", "b","c","d","e"]    #预测

def detect_face(img):
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    face_cascade = cv2.CascadeClassifier('../cv-class/lbpcascade_frontalface.xml')
    faces = face_cascade.detectMultiScale(gray, scaleFactor=1.2, minNeighbors=1)
    if (len(faces) == 0):
        return None, None
    (x, y, w, h) = faces[0]
    return gray[y:y + w, x:x + h], faces[0]

def prepare_training_data(data_folder_path):
    dirs = os.listdir(data_folder_path)
    faces = []
    labels = []
    for dir_name in dirs:
        if not dir_name.startswith("s"):
            continue;
        label = int(dir_name.replace("s", ""))

        subject_dir_path = data_folder_path + "/" + dir_name
        subject_images_names = os.listdir(subject_dir_path)

        for image_name in subject_images_names:
            if image_name.startswith("."):
                continue;
            image_path = subject_dir_path + "/" + image_name
            image = cv2.imread(image_path)
            cv2.imshow("Training on image...", cv2.resize(image, (400, 500)))
            cv2.waitKey(100)

            face, rect = detect_face(image)
            if face is not None:
                faces.append(face)
                labels.append(label)
    cv2.waitKey(1)
    cv2.destroyAllWindows()
    return faces, labels

print("Preparing data...")
faces, labels = prepare_training_data("training-data")  #调用训练库
print("Total faces: ", len(faces))
print("Total labels: ", len(labels))

face_recognizer = cv2.face.createLBPHFaceRecognizer()
face_recognizer.train(faces, np.array(labels))          #开始训练模型
face_recognizer.save('training-data/trainner.yml')