python行业缺口_opencv-python 缺口识别

一、cv函数

1、imread：读取图片

imread(image_path, flag):

images_path:图片路径，找不到不报错

flag:

1/cv2.IMREAD_COLOR:彩色图片,图片透明性会被忽略,默认参数

0/cv2.IMREAD_GRAYSCALE:灰色图片

-1/cv2.IMREAD_UNCHANGED:包括其alpha通道

2、imwrite

imwrite(img_path_name,img)

img_path_name:保存的文件名

img：文件对象

3、cvtColor

cvtColor(img,code)

img: 图像对象

code：

cv2.COLOR_RGB2GRAY: RGB转换到灰度模式

cv2.COLOR_RGB2HSV： RGB转换到HSV模式（hue,saturation,Value）

4、matchTemplate

matchTemplate(img_path, bg_path, cv2.TM_CCOEFF_NORMED)

img_path:对比图片

bg_path：背景图片

cv2.TM_CCOEFF_NORMED

```

# encoding=utf8

import cv2

import numpyas np

def show(name):

cv2.imshow('Show', name)

cv2.waitKey(0)

cv2.destroyAllWindows()

def main():

otemp ='./images/tb.png'

oblk ='./images/bg.jpg'

target = cv2.imread(otemp, 0)

template = cv2.imread(oblk, 0)# 读取到两个图片，进行灰值化处理

w, h = target.shape[::-1]

aa = target.shape

print(aa)

print(w, h)

temp ='./images/temp.jpg'

targ ='./images/targ.jpg'

cv2.imwrite(temp, template)

cv2.imwrite(targ, target)# 处理后进行保存

target = cv2.imread(targ)

target = cv2.cvtColor(target, cv2.COLOR_BGR2GRAY)# 转化到灰度

target =abs(255 - target)# 返回绝对值

cv2.imwrite(targ, target)# 重新写入

target = cv2.imread(targ)

template = cv2.imread(temp)

result = cv2.matchTemplate(target, template, cv2.TM_CCOEFF_NORMED)# 进行匹配

x, y = np.unravel_index(result.argmax(), result.shape)# 通过np转化为数值，就是坐标

print(y, x)

# 展示圈出来的区域

cv2.rectangle(template, (y, x), (y + w, x + h), (7, 249, 151), 2)

show(template)

return y, x

if __name__ =='__main__':

a, b = main()

```