所需要的库

face_recognition 1.2.3

dlib 19.8.1

---

首先加载图片

Gal_Gadot_image = face_recognition.load_image_file("E:/Gal_Gadot_1.jpg")

Scarlett_Johansson_image = face_recognition.load_image_file("E:/Scarlett_Johansson.jpg")

unknown_image = face_recognition.load_image_file("E:/Gal_Gadot_2.jpg")

---

加载的图像为三维矩阵

对图片进行特征提取

Gal_Gadot_face_encoding = face_recognition.face_encodings(Gal_Gadot_image)[0]

Scarlett_Johansson_face_encoding = face_recognition.face_encodings(Scarlett_Johansson_image)[0]

unknown_face_encoding = face_recognition.face_encodings(unknown_image)[0]

face_recognition.face_encodings会返回图片中的所有的人脸的128位向量

因为是单人照片所以只取第一个元素

Gal_Gadot_face_encoding = face_recognition.face_encodings(Gal_Gadot_image)[0]

把人脸的特征向量保存起来

known_faces = [

Gal_Gadot_face_encoding,

Scarlett_Johansson_face_encoding]

---

把需要识别的图片与已经识别的图片进行对比

results = face_recognition.compare_faces(known_faces, unknown_face_encoding)

---

输出的结果

[True, False]

#结果识别正确

---

该图片用到的照片为盖尔加朵和斯嘉丽约翰逊

哈哈哈