shape\_predictor\_68\_face\_landmarks.dat 下載 : https://blog.csdn.net/hotlinhao/article/details/119415047 ## shape\_predictor\_68\_face\_landmarks.dat 下載： https://codeleading.com/article/8170867908/#google_vignette ``` from imutils import face_utils import dlib import imutils import cv2 detector = dlib.get_frontal_face_detector() predictor = dlib.shape_predictor("F:\\postgraduate\\Project\\FaceRecognitionBasedOnHogSVM\\shape_predictor_68_face_landmarks.dat") image = cv2.imread("example_08.jpg") image = imutils.resize(image, width=500) gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) rects = detector(gray, 1) # enumerate()方法用于將一個可遍歷的數據對象(列表、元組、字典)組合 # 為一個索引序列，同時列出 數據下標 和 數據 ，一般用在for循環中 for(i, rect) in enumerate(rects): shape = predictor(gray, rect) # 標記人臉中的68個landmark點 shape = face_utils.shape_to_np(shape) # shape轉換成68個坐標點矩陣 (x, y, w, h) = face_utils.rect_to_bb(rect) # 返回人臉框的左上角坐標和矩形框的尺寸 cv2.rectangle(image, (x, y), (x + w, y + h), (0, 255, 0), 2) cv2.putText(image, "Face #{}".format(i + 1), (x - 10, y - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2) landmarksNum = 0; for (x, y) in shape: cv2.circle(image, (x, y), 2, (0, 0, 255), -1) # cv2.putText(image, "{}".format(landmarksNum), (x, y), # cv2.FONT_HERSHEY_SIMPLEX, 0.2, (255, 0, 0), 1) # landmarksNum = landmarksNum + 1; landmarksNum = 0; cv2.imshow("Output", image) cv2.waitKey(0) ``` demo2 ``` import dlib import numpy as np import cv2 # 加載人臉檢測器 detector = dlib.get_frontal_face_detector() # 加載人臉關鍵點檢測器 predictor = dlib.shape_predictor('shape_predictor_68_face_landmarks.dat') # 加載人臉識別模型 sp = dlib.face_recognition_model_v1('dlib_face_recognition_resnet_model_v1.dat') # 加載圖像 img = cv2.imread('test.jpg') # 進行人臉檢測 dets = detector(img, 1) # 對于每個人臉，進行關鍵點檢測和人臉識別 for k, d in enumerate(dets): shape = predictor(img, d) face_descriptor = sp.compute_face_descriptor(img, shape) print('Face {}: {}'.format(k, face_descriptor)) ```