To identify a specific image, the key is to have a signature file to identify the object, OpenCV has built-in
face recognition profile, we simply use the OpenCV CascadeClassifier class can be identified.
Creating Cas cadeClas s ifier object syntax is:
import cv2 faceCascade = cv2.CascadeClassifier("E:\\haarcascade_frontalface_default.xml")
Can then be identified by facial recognition target detectMultiScale method, the syntax is:
For example, to identify the minimum set area (10, 10), the false detection rate is 5 , the normal pattern recognition, to recognize the
image as Image, and the recognition result stored in the variable to the faces:
Identifying the position of the face
to identify the face position in the image, and displays the number in the lower left corner of the face recognition.
import cv2 casc_path = "E:\\haarcascade_frontalface_default.xml" faceCascade = cv2.CascadeClassifier(casc_path) imagename = cv2.imread("F:\\pythonBase\\pythonex\\ch10\\media\\person1.jpg") faces = faceCascade.detectMultiScale(imagename, scaleFactor=1.1, minNeighbors=5, minSize=(30,30), flags = cv2.CASCADE_SCALE_IMAGE) #imagename.shape[0]:图片高度，imagename.shape[1]:图片宽度 cv2.rectangle(imagename, (10,imagename.shape[0]-20), (110,imagename.shape[0]), (0,0,0), -1) cv2.putText(imagename,"Find " + str(len(faces)) + " face!", (10,imagename.shape[0]-5), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255,255,255), 2) for (x,y,w,h) in faces: cv2.rectangle(imagename,(x,y),(x+w, y+h),(128,255,0),2) cv2.namedWindow("facedetect") cv2.imshow("facedetect", imagename) cv2.waitKey(0) cv2.destroyWindow("facedetect")
detectMultiScale 方法可识别图片中的多个人脸，把图
片文件更改为 person3.jpg 或 person8.jpg 进行多个人脸的识别 。
import cv2 casc_path = "E:\\haarcascade_frontalface_default.xml" faceCascade = cv2.CascadeClassifier(casc_path) imagename = cv2.imread("F:\\pythonBase\\pythonex\\ch10\\media\\person3.jpg") faces = faceCascade.detectMultiScale(imagename, scaleFactor=1.1, minNeighbors=5, minSize=(30,30), flags = cv2.CASCADE_SCALE_IMAGE) #imagename.shape[0]:图片高度，imagename.shape[1]:图片宽度 cv2.rectangle(imagename, (10,imagename.shape[0]-20), (110,imagename.shape[0]), (0,0,0), -1) cv2.putText(imagename,"Find " + str(len(faces)) + " face!", (10,imagename.shape[0]-5), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255,255,255), 2) for (x,y,w,h) in faces: cv2.rectangle(imagename,(x,y),(x+w, y+h),(128,255,0),2) cv2.namedWindow("facedetect") cv2.imshow("facedetect", imagename) cv2.waitKey(0) cv2.destroyWindow("facedetect")
import cv2 casc_path = "E:\\haarcascade_frontalface_default.xml" faceCascade = cv2.CascadeClassifier(casc_path) imagename = cv2.imread("F:\\pythonBase\\pythonex\\ch10\\media\\person8.jpg") faces = faceCascade.detectMultiScale(imagename, scaleFactor=1.1, minNeighbors=5, minSize=(30,30), flags = cv2.CASCADE_SCALE_IMAGE) #imagename.shape[0]:图片高度，imagename.shape[1]:图片宽度 cv2.rectangle(imagename, (10,imagename.shape[0]-20), (110,imagename.shape[0]), (0,0,0), -1) cv2.putText(imagename,"Find " + str(len(faces)) + " face!", (10,imagename.shape[0]-5), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255,255,255), 2) for (x,y,w,h) in faces: cv2.rectangle(imagename,(x,y),(x+w, y+h),(128,255,0),2) cv2.namedWindow("facedetect") cv2.imshow("facedetect", imagename) cv2.waitKey(0) cv2.destroyWindow("facedetect")
