[人工智能]opencv学习_pthon版opencv基础使用

opencv使用

绘制图片

#!/usr/bin/env python3
# -*- coding: utf-8 -*-


def main(args):
	import cv2
	import numpy
	
	mat = numpy.zeros((300,400),dtype = numpy.uint8)
	
	mat = cv2.cvtColor(mat,cv2.COLOR_GRAY2BGR)
	cv2.imshow('main',mat)
	cv2.waitKey()
	
	return 0

if __name__ == '__main__':
	import sys
	sys.exit(main(sys.argv))

显示图片

#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
#  showpic.py
#  
import cv2

def main(args):
	img = cv2.imread('/home/pi/Pictures/test.jpg', cv2.IMREAD_COLOR)
	cv2.imshow('image', img)
	cv2.waitKey()

if __name__ == '__main__':
	import sys
	sys.exit(main(sys.argv))

人脸识别

import cv2

img = cv2.imread("model.png")  # 读取人脸图像
# 加载识别人脸的级联分类器
faceCascade = cv2.CascadeClassifier("cascades/haarcascade_frontalface_default.xml")
faces = faceCascade.detectMultiScale(img, 1.3)  # 识别出所有人脸
for (x, y, w, h) in faces:  # 遍历所有人脸的区域
    cv2.rectangle(img, (x, y), (x + w, y + h), (0, 0, 255), 5)  # 在图像中人脸的位置绘制方框
cv2.imshow("img", img)  # 显示最终处理的效果
cv2.waitKey()  # 按下任何键盘按键后
cv2.destroyAllWindows()  # 释放所有窗体

人脸识别训练

# Eigenfaces人脸识别器
# recognizer = cv2.face.EigenFaceRecognizer_create(num_components, threshold) #均使用默认值
# recognizer.train(src,labels) 此处的src为待训练图像组
# label,confidence = recognizer.predict(src) 此处的src为待测试图像

import cv2
import numpy

photos = list()
labels = list()

photos.append(cv2.imread("face/summer1.png",0)) #cv2.imread(filename,flags) flags默认为1显示彩色,设置为0则显示灰度图
labels.append(0)
photos.append(cv2.imread("face/summer2.png",0))
labels.append(0)
photos.append(cv2.imread("face/summer3.png",0))
labels.append(0)

photos.append(cv2.imread("face/Elvis1.png",0))
labels.append(1)
photos.append(cv2.imread("face/Elvis2.png",0))
labels.append(1)
photos.append(cv2.imread("face/Elvis3.png",0))
labels.append(1)

test = cv2.imread("face/summer4.png",0)
names = {"0":"Summer","1":"Elvis"}

recognizer = cv2.face.EigenFaceRecognizer_create()
recognizer.train(photos,numpy.array(labels))
label,confidence = recognizer.predict(test)

print("confidence = " + str(confidence))
print(names[str(label)])