效果
流程如下:先选择canny的两个阈值,确定好需要进行模糊的边缘,再选择模糊程度
原理
canny边缘检测
Canny边缘检测算法可以分为以下5个步骤:
-
使用高斯滤波器,以平滑图像,滤除噪声。
-
计算图像中每个像素点的梯度强度和方向。
-
应用非极大值(Non-Maximum Suppression)抑制,以消除边缘检测带来的杂散响应。
-
应用双阈值(Double-Threshold)检测来确定真实的和潜在的边缘。
-
通过抑制孤立的弱边缘最终完成边缘检测。
代码
import cv2
t1 = 0
t2 = 0
def Preview(img_path,ts1,ts2):
img = cv2.imread(img_path)
img_copy = img.copy()
imgCanny = cv2.Canny(img,ts1,ts2)
g = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 5))
g2 = cv2.getStructuringElement(cv2.MORPH_RECT, (7, 7))
img_dilate = cv2.dilate(imgCanny, g)
img_dilate2 = cv2.dilate(imgCanny, g2)
shape = img_dilate.shape
for i in range(shape[0]):
for j in range(shape[1]):
if img_dilate2[i,j] == 0:
img[i,j] = [0,0,0]
cv2.namedWindow("Preview",cv2.WINDOW_NORMAL)
cv2.createTrackbar("kerner_size","Preview",0,7,nothing)
while(1):
size = cv2.getTrackbarPos("kerner_size","Preview")
dst = cv2.GaussianBlur(img,(2*size+1,2*size+1),0,0,cv2.BORDER_DEFAULT)
for i in range(shape[0]):
for j in range(shape[1]):
if img_dilate[i,j] != 0:
img_copy[i,j] =dst[i,j]
cv2.imshow("Result",img_copy)
k = cv2.waitKey(1) & 0xff
if k == 27:
break
cv2.destroyAllWindows()
def Choose_Thre(img_path):
global t1,t2
img = cv2.imread(img_path)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
cv2.namedWindow("Preview",cv2.WINDOW_NORMAL)
def nothing(WEV):
pass
cv2.createTrackbar("Threshold1","Preview",0,255,nothing)
cv2.createTrackbar("Threshold2","Preview",0,255,nothing)
while(1):
t1 = cv2.getTrackbarPos("Threshold1","Preview")
t2 = cv2.getTrackbarPos("Threshold2","Preview")
detected_edges = cv2.GaussianBlur(gray,(3,3),0)
detected_edges = cv2.Canny(detected_edges,t1,t2,apertureSize = 3)
dst = cv2.bitwise_and(img,img,mask = detected_edges)
cv2.imshow('Edge',dst)
k = cv2.waitKey(1)&0xFF
if k==27:
break
cv2.destroyAllWindows()
def nothing(wte):
pass
#按esc退出 直接关闭无效
img_path = '1.jpg'
Choose_Thre(img_path)
print(t1,t2)
Preview(img_path,t1,t2)