-
Notifications
You must be signed in to change notification settings - Fork 0
/
watershed_4.py
98 lines (90 loc) · 3.54 KB
/
watershed_4.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
'''
Version: 1.0
Author: xiawei
Date: 2022-12-27 10:26:51
LastEditors: xiawei
LastEditTime: 2022-12-31 09:47:08
Description:
灰度,高斯，然后二值，找连通域，按面积筛一筛
'''
import numpy as np
import cv2 as cv2
# 读取,灰度,高斯,二值化
img = cv2.imread('./1.png')
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
blurred = cv2.GaussianBlur(gray_img, (11, 11), 0)
# blurred = cv2.bilateralFilter(gray_img, 3, 13, 23, 10)
threshold = cv2.threshold(blurred, 205, 255,
cv2.THRESH_BINARY)[1]
# 对二值化分析连通域
totalLabels, label_ids, stats, centroid = cv2.connectedComponentsWithStats(threshold,
4,
cv2.CV_32S)
mark = img.copy()
# mark[label_ids == 1] = (255, 0, 0) # 连通域/背景（蓝）
# mark[label_ids == 0] = (0, 255, 0) # 不确定区域（绿）
# mark[label_ids > 1] = (0, 0, 255) # 前景/种子（红）
# mark[label_ids == -1] = (255, 0, 0) # 边界（绿）
# cv2.namedWindow('Markers', cv2.WINDOW_NORMAL)
# cv2.imshow('Markers', mark)
# 定义输出矩阵
output = np.zeros(gray_img.shape, dtype="uint8")
print('totalLabels', totalLabels)
for i in range(1, totalLabels):
# 提取当前标签的连通分量统计信息,这里提取面积
x = stats[i, cv2.CC_STAT_LEFT]
y = stats[i, cv2.CC_STAT_TOP]
w = stats[i, cv2.CC_STAT_WIDTH]
h = stats[i, cv2.CC_STAT_HEIGHT]
area = stats[i, cv2.CC_STAT_AREA]
(cX, cY) = centroid[i]
print('w', w)
print('h', h)
print('area', area)
# 确保宽高以及面积既不太大也不太小
keepWidth = w > 3000 and w < 3665
keepHeight = h > 2000 and h < 2750
keepArea = area > 3320 and area < 9460540
# keepWidth = w == 3664
# keepHeight = h == 2748
# keepArea = area == 9471470
# 随着高斯核变大 区域面积变大9512393,因为模糊了——9550580
if (area > 1000):
print('come in')
componentMask = (label_ids == i).astype("uint8") * 255
output = cv2.bitwise_or(output, componentMask)
cv2.namedWindow('output', cv2.WINDOW_NORMAL)
cv2.imshow('output', output)
"""if all((keepWidth, keepHeight, keepArea)):
print('21')
componentMask = (label_ids == i).astype("uint8") * 255
output = cv2.bitwise_or(output, componentMask)
cv2.namedWindow('output', cv2.WINDOW_NORMAL)
cv2.imshow('output', output)"""
img22 = img.copy()
canny_img = cv2.Canny(output, 150, 300)
# 查找轮廓api，原始图像，查找轮廓方式RETR_EXTERNAL为之查找外围，List从里到外从右到左，CCOMP有层级关系没搞懂
# 从大到小从右到左,这里的contours为根据CHAIN_APPROX_SIMPLE设定的点。
# 返回轮廓个数和坐标值
area2 = []
contours, _ = cv2.findContours(
canny_img, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
for k in range(len(contours)):
area_k = cv2.contourArea(contours[k])
if (area_k > 2000):
area2.append(k)
# max_idx = np.argmax(np.array(area2))
# 最大像素为212129.5
# print("area2222", cv2.contourArea(contours[max_idx]))
# 绘制轮廓，-1为绘制所有轮廓，颜色，线宽。
# print('masker', _)
for j in area2:
img2 = cv2.drawContours(img, contours, j, (0, 0, 255), 5)
cv2.namedWindow('img2', cv2.WINDOW_NORMAL)
cv2.imshow('img2', img2)
cv2.namedWindow('canny_img', cv2.WINDOW_NORMAL)
cv2.imshow('canny_img', canny_img)
cv2.waitKey()
cv2.namedWindow('img', cv2.WINDOW_NORMAL)
cv2.imshow('img', img22)
cv2.waitKey(0)