python 用opencv调用训练好的模型进行识别的方法_Python

python 用opencv调用训练好的模型进行识别的方法

2021-04-26 00:27interstellar-ai Python

今天小编就为大家分享一篇python 用opencv调用训练好的模型进行识别的方法，具有很好的参考价值，希望对大家有所帮助。一起跟随小编过来看看吧

此程序为先调用opencv自带的人脸检测模型,检测到人脸后,再调用我自己训练好的模型去识别人脸,使用时更改模型地址即可

									#!usr/bin/env python

									import cv2

									font=cv2.FONT_HERSHEY_SIMPLEX

									cascade1 = cv2.CascadeClassifier("D:\\opencv249\\opencv\\sources\\data\\haarcascades\\haarcascade_frontalface_alt_tree.xml")

									cascade2 = cv2.CascadeClassifier("D:\\opencv249\\opencv\\sources\\data\\haarcascades\\xml.xml")

									cap = cv2.VideoCapture(0)

									while True:

									 ret,frame = cap.read()

									 gray = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)

									 rect = cascade1.detectMultiScale(gray,scaleFactor=1.3,minNeighbors=9,minSize=(50,50),flags = cv2.cv.CV_HAAR_SCALE_IMAGE)

									 if not rect is (): 

									  for x,y,z,w in rect:

									   roiImg = gray[y:y+w,x:x+z]

									   rect1 = cascade2.detectMultiScale(roiImg,scaleFactor=1.3,minNeighbors=9,minSize=(50,50),flags = cv2.cv.CV_HAAR_SCALE_IMAGE)

									   if not rect1 is ():

									    for (a,b,c,d) in rect1:

									     print "rect",rect1

									     cv2.putText(frame,'Chenym',(x,y), font, 2,(0,0,255),2)

									     cv2.rectangle(frame,(x+a,y+b),(x+a+c,y+b+d),(0,0,255),2)

									 cv2.imshow('frame',frame)

									 if cv2.waitKey(1) &0xFF == ord('q'):

									  break

									cap.release()

									cv2.destroyAllWindows()