人脸关键点检测——dlib

1、图片识别

 人脸关键点检测——dlib_第1张图片人脸关键点检测——dlib_第2张图片

 

注意点:

1. dlib.get_frontal_face_detector( ) 获取人脸检测器

2. dlib.shape_predictor( ) 预测人脸关键点

人脸关键点模型,下载地址:

# 1 加入库
import cv2
import matplotlib.pyplot as plt
import dlib

# 2 读取一张图片
image = cv2.imread("Tom.jpeg")

# 3 调用人脸检测器
detector = dlib.get_frontal_face_detector()

# 4 加载预测关键点模型(68个关键点)
# 人脸关键点模型,下载地址:
# http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2.
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")

# 5 灰度转换
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

# 6 人脸检测
faces = detector(gray, 1)

# 7 循环,遍历每一张人脸,给人脸绘制矩形框和关键点
for face in faces: #(x, y, w, h)
    # 8 绘制矩形框
    cv2.rectangle(image, (face.left(), face.top()), (face.right(), face.bottom()), (0,255,0), 2)

    # 9 预测关键点
    shape = predictor(image, face)

    # 10 获取到关键点坐标
    for pt in shape.parts():
        # 获取横纵坐标
        pt_position = (pt.x, pt.y)
        # 11 绘制关键点坐标
        cv2.circle(image, pt_position, 1, (255, 0, 0), -1)# -1填充,2表示大小

# 12 显示整个效果图
plt.imshow(image)
plt.axis("off")
plt.show()



2、电脑摄像头识别

# 1 加入库
import cv2
import dlib

# 2 打开摄像头
capture = cv2.VideoCapture(0)

# 3 获取人脸检测器
detector = dlib.get_frontal_face_detector()

# 4 获取人脸关键点检测模型
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")

while True:
    # 5 读取视频流
    ret, frame = capture.read()
    # 6 灰度转换
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    # 7 人脸检测
    faces = detector(gray, 1)
    # 8 绘制每张人脸的矩形框和关键点
    for face in faces:
        # 8.1 绘制矩形框
        cv2.rectangle(frame, (face.left(), face.top()), (face.right(), face.bottom()), (0,255,0), 3)
        # 8.2 检测到关键点
        shape = predictor(gray, face)  #68个关键点
        # 8.3 获取关键点的坐标
        for pt in shape.parts():
            # 每个点的坐标
            pt_position = (pt.x, pt.y)
            # 8.4 绘制关键点
            cv2.circle(frame, pt_position, 3, (255,0,0), -1)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
    # 9 显示效果
    cv2.imshow("face detection landmark", frame)
capture.release()
cv2.destroyAllWindows()

你可能感兴趣的:(计算机视觉,opencv,人工智能,人脸识别)