else:
cv2.rectangle(image, (x - 5, y - 15), (x + w, y + h + 10), (255, 255, 255), 20)
elif k == 3:
if w < bounding_list[3][2]:
w = bounding_list[3][2]
cv2.rectangle(image, (x - 5, y - 5), (x + w, y + h + 10), (255, 255, 255), 25)
else:
cv2.rectangle(image, (x - 5, y - 5), (x + w, y + h + 10), (255, 255, 255), 25)
elif k == 4:
if w < bounding_list[3][2]:
w = bounding_list[3][2]
if x > bounding_list[3][0]:
x = bounding_list[3][0]
cv2.rectangle(image, (x - 5, y - 15), (x + w, y + h + 14), (255, 255, 255), 25)
else:
cv2.rectangle(image, (x - 5, y - 15), (x + w, y + h + 14), (255, 255, 255), 25)
cv2.line(image, (x, y + h - 10), (x + w, y + h - 10), (255, 255, 255), 25)
return image
if __name__ == '__main__':
images = ['C:/Users/Administrator/Desktop/1_ELwangkang/ASIC_PD_05_03_21_022646.jpg','C:/Users/Administrator/Desktop/1_ELwangkang/ASIC_PD_05_03_21_022450.jpg']
original = cv2.imread(images[1])
image_show('测试', original)
image = original
add_rectangle_image = draw_rectangle(image, original)
image_show('add_rectangle_image', add_rectangle_image)
:param standard_light_coef: 目标亮度系数.
:param standard_contrast_coef: 目标对比度系数.
:return: over_picture,
"""
needed_image = cv2.imread(picture)
needed_image_light_coef = light_contrast_coef(needed_image)[0]
needed_image_contrast_coef = light_contrast_coef(needed_image)[
1] # 待调图片的亮度信息
needed_image_contrast = needed_image - needed_image_light_coef # 待调图片的对比度信息 数组（广播）
m = standard_light_coef / needed_image_light_coef # 亮度系数
n = standard_contrast_coef / needed_image_contrast_coef # 对比度系数
over_picture = needed_image_light_coef * m + needed_image_contrast * n # 修改亮度和对比度
# new_array_picture = new_array_picture/255
# 保存到本地注释掉，在batch_deal_el.py中保存到本地，便于批处理
#cv2.imwrite('D://Documents//Desktop//over_picture.jpg', over_picture)
return over_picture
if __name__ == '__main__':
from tools import image_show
# 图片img的亮度系数、对比度系数
img = cv2.imread('D://Documents//Desktop//sxdhd.jpg')
standard_light_coef = light_contrast_coef(img)[0]
standard_contrast_coef = light_contrast_coef(img)[1]
over_picture = adjust_image_info('D://Documents//Desktop//szy2.jpg',
standard_light_coef,
standard_contrast_coef)
image_show('over picture', over_picture)
db="ak_zuul",
charset="utf8")
cur = conn.cursor()
sql = "INSERT INTO aiko_elprint (original_el_path, constract_el_path,rectangle_el_path,tzqt_el_path,dytzqt_el_path,dytz_el_path) VALUES ( '%s', '%s','%s', '%s', '%s','%s')"
data = (original_need_deal_el_path, constract_el_path,
rectangle_el_path, tzqt_el_path, dytzqt_el_path, dytz_el_path)
cur.execute(sql % data)
conn.commit()
# print('边缘矩形图像已保存到本地')
if __name__ == '__main__':
from tools import image_show
over_picture_path = 'F://2017//12//ELTD//Renamepath2//101985314740.jpg'
# original_image = cv2.imread('D://Documents//Desktop//over_picture.jpg') # 改变亮度、对比度后的图片image1
image = cv2.imread('F://2017//test//Rectangle//Rectangle_101985314740.jpg'
) # 在image1的基础上增加矩形边缘、白色格栅线的图片image2
# 图片img的二值图和轮廓等。
thresh = Contours_feature(image)[0]
transformed_image = transform_form(thresh, 'closing')
image_show('transformed_image', transformed_image) # 展示形态学转换后的二值图
new_image, new_contours, new_hierarchy = cv2.findContours(
transformed_image, cv2.RETR_EXTERNAL,
cv2.CHAIN_APPROX_SIMPLE) # image2的二值图经过形态学转化后的轮廓等提取
bounding_show(new_contours, over_picture_path,
number_id=None) # 在image1 中展示故障特征
# save_bounding_picture(new_contours, over_picture_path) # 根据轮廓，提取故障特征保存在本地
"""
imgray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
ret, thresh = cv2.threshold(imgray, 10, 255, cv2.THRESH_BINARY)
# 反色，即对二值图每个像素取反
result = cv2.bitwise_not(thresh)
dilation_kernel = np.ones((10, 10), np.uint8)
dilation = cv2.dilate(result, dilation_kernel, iterations=2)
erode_kernel = np.ones((5, 5), np.uint8)
closing_image = cv2.erode(dilation, erode_kernel, iterations=1)
new_image, new_contours, new_hierarchy = cv2.findContours(
closing_image, cv2.RETR_TREE,
cv2.CHAIN_APPROX_SIMPLE) # image2的二值图经过形态学转化后的轮廓等
for i in range(len(new_contours)):
j = new_contours[i]
x, y, w, h = cv2.boundingRect(j)
cv2.rectangle(image, (x - 10, y - 10), (x + w + 8, y + h + 5),
(255, 255, 255),
17) # 不注释的话会使之后的img变量表示的图片上都是绿色的线框存在，影响图片原有信息。
# 保存到本地注释掉，在batch_deal_el.py中保存到本地，便于批处理
# cv2.imwrite('D://Documents//Desktop//add_rectangle_image.jpg', ee)
return image
if __name__ == '__main__':
from tools import image_show
image = cv2.imread('F:/2017/12/ELTD/Constract/Constract_101910111517.jpg')
add_rectangle_image = draw_rectangle(image)
image_show('add_rectangle_image', add_rectangle_image)