人脸识别相关代码

人脸识别相关代码

这个是网上别人做的视频的学习地址：
https://www.bilibili.com/video/BV1zc41197BA/?p=8&vd_source=e9167c654bb4523338a765358a8ac2af
手敲了一遍：

总结

我只能说自己训练的人脸效果实在是不好，还是没搞懂为啥别人的视频效果确实不错，或许是训练集问题

人脸识别

其中haarcascade_frontalface_default.xml，这个下载opencv以后就会有的文件，然后将其添加到项目中

import cv2 as cv


def detect():
    gray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)
    # 使用绝对路径加载级联分类器文件
    face_detector = cv.CascadeClassifier('./haarcascades/haarcascade_frontalface_default.xml')
    faces = face_detector.detectMultiScale(gray)

    for (x, y, w, h) in faces:
        cv.rectangle(img, (x, y), (x + w, y + h), (0, 0, 255), 2)
    cv.imshow('result', img)
    cv.waitKey(0)


img = cv.imread('1.jpeg')
if __name__ == '__main__':
    detect()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

检测视频中的人脸

import cv2 as cv

def detect(img):
    gray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)
    # 使用绝对路径加载级联分类器文件
    face_detector = cv.CascadeClassifier('./haarcascades/haarcascade_frontalface_default.xml')
    faces = face_detector.detectMultiScale(gray)

    for (x, y, w, h) in faces:
        cv.rectangle(img, (x, y), (x + w, y + h), (0, 0, 255), 2)
    cv.imshow('result', img)
    cv.waitKey(0)

def main():
    cap = cv.VideoCapture('video.mp4')
    while True:
        flag, frame = cap.read()
        if not flag:
            break
        detect(frame)
        if ord('q') == cv.waitKey(10):
            break
    cv.destroyAllWindows()
    cap.release()
if __name__ == '__main__':
    main()

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

训练数据

import os

import cv2 as cv
import numpy as np
from PIL import Image


def getImageAndLabels(path):
    # 使用绝对路径加载级联分类器文件
    face_detector = cv.CascadeClassifier('./haarcascades/haarcascade_frontalface_default.xml')
    faceSamples = []
    ids = []
    imagePaths = [os.path.join(path, f) for f in os.listdir(path)]
    for imagePaths in imagePaths:
        PIL_img = Image.open(imagePaths).convert('L')
        # 图像转为数组
        img_numpy = np.array(PIL_img, 'uint8')
        faces = face_detector.detectMultiScale(img_numpy)
        # 获取每张图像的id
        id = int(os.path.split(imagePaths)[1].split('.')[0])
        for x, y, w, h in faces:
            faceSamples.append(img_numpy[y:y + h, x:x + w])
            ids.append(id)
    return faceSamples, ids


if __name__ == '__main__':
    path = './data/jm/'
    faces, ids = getImageAndLabels(path)
    recognizer = cv.face.LBPHFaceRecognizer_create()
    recognizer.train(faces, np.array(ids))
    # 保存文件
    recognizer.write('trainer.yml')

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

置信区间

import cv2 as cv

def detect(img):
    gray = cv.cvtColor(img, cv.COLOR_BGR2GRAY)
    # 使用绝对路径加载级联分类器文件
    face_detector = cv.CascadeClassifier('./haarcascades/haarcascade_frontalface_default.xml')
    faces = face_detector.detectMultiScale(gray)

    recognizer = cv.face.LBPHFaceRecognizer_create()
    recognizer.read('trainer.yml')

    for (x, y, w, h) in faces:
        roi_gray = gray[y:y + h, x:x + w]
        label, confidence = recognizer.predict(roi_gray)

        # 打印预测标签和置信度
        print(f"Predicted Label: {label}, Confidence: {confidence}")

        # 绘制矩形框
        cv.rectangle(img, (x, y), (x + w, y + h), (100, 255, 0), 2)

        # 计算文本位置
        text_x = x + int(w / 2) - 50
        text_y = y - 20

        # 绘制标签和置信度信息
        label_text = f"Label: {label}, Confidence: {confidence:.2f}"
        cv.putText(img, label_text, (text_x, text_y), cv.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)

    cv.imshow('result', img)
    cv.waitKey(0)
    cv.destroyAllWindows()

if __name__ == '__main__':
    path = './data/test/'
    detect(cv.imread(path + '9.jpg'))

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37