基于图像处理的皱纹检测算法。

https://github.com/bulingda/Wrinkles_detection/blob/master/Wrinkle.py

基于RCNN 毛孔检测

https://github.com/jack16888/Faster-RCNN-for-pore-detection

	from PIL import Image
	import cv2
	import time
	import numpy as np
	from skimage.filters import frangi, gabor
	from skimage import measure, morphology
	from Partition import practice14 as pa
	#
	# class UnionFindSet(object):
	#
	# def __init__(self, m):
	# # m, n = len(grid), len(grid[0])
	# self.roots = [i for i in range(m)]
	# self.rank = [0 for i in range(m)]
	# self.count = m
	#
	# for i in range(m):
	# self.roots[i] = i
	#
	# def find(self, member):
	# tmp = []
	# while member != self.roots[member]:
	# tmp.append(member)
	# member = self.roots[member]
	# for root in tmp:
	# self.roots[root] = member
	# return member
	#
	# def union(self, p, q):
	# parentP = self.find(p)
	# parentQ = self.find(q)
	# if parentP != parentQ:
	# if self.rank[parentP] > self.rank[parentQ]:
	# self.roots[parentQ] = parentP
	# elif self.rank[parentP] < self.rank[parentQ]:
	# self.roots[parentP] = parentQ
	# else:
	# self.roots[parentQ] = parentP
	# self.rank[parentP] -= 1
	# self.count -= 1
	#
	#
	# class Solution(object):
	# def countComponents(self, n, edges):
	# """
	# :type n: int
	# :type edges: List[List[int]]
	# :rtype: int
	# """
	# ufs = UnionFindSet(n)
	# # print ufs.roots
	# for edge in edges:
	# start, end = edge[0], edge[1]
	# ufs.union(start, end)
	#
	# return ufs.count
	#
	#
	# # def connected_component(thresh_A):
	# # thresh_A_copy = thresh_A.copy() # 复制thresh_A到thresh_A_copy
	# # thresh_B = np.zeros(thresh_A.shape).astype(int) # thresh_B大小与A相同，像素值为0
	# #
	# # kernel = cv2.getStructuringElement(cv2.MORPH_CROSS, (3, 3)) # 3×3结构元
	# #
	# # count = [] # 为了记录连通分量中的像素个数
	# # xy_count = []
	# # # 循环，直到thresh_A_copy中的像素值全部为0
	# # while thresh_A_copy.any():
	# # Xa_copy, Ya_copy = np.where(thresh_A_copy > 0) # thresh_A_copy中值为255的像素的坐标
	# # thresh_B[Xa_copy[0]][Ya_copy[0]] = 1 # 选取第一个点，并将thresh_B中对应像素值改为255
	# #
	# # # 连通分量算法，先对thresh_B进行膨胀，再和thresh_A执行and操作（取交集）
	# # for i in range(thresh_A.shape[0]):
	# # dilation_B = cv2.dilate(thresh_B, kernel, iterations=1)
	# # print(type(thresh_A), type(dilation_B.astype(int)))
	# # thresh_B = cv2.bitwise_and(np.uint8(thresh_A), dilation_B.astype(int))
	# #
	# # # 取thresh_B值为255的像素坐标，并将thresh_A_copy中对应坐标像素值变为0
	# # Xb, Yb = np.where(thresh_B > 0)
	# # thresh_A_copy[Xb, Yb] = 0
	# # xy_count.append(np.c_[Xb, Yb])
	# # # 显示连通分量及其包含像素数量
	# # count.append(len(Xb))
	# # if len(count) == 0:
	# # print("无连通分量")
	# # if len(count) == 1:
	# # print("第1个连通分量为{}".format(count[0]))
	# # print(xy_count[0], "2")
	# # ecc = measure.regionprops(xy_count[0])
	# # print(ecc[0].eccentricity)
	# # if len(count) >= 2:
	# # if (count[-1] - count[-2]) < 2:
	# # continue
	# # print("第{}个连通分量为{}".format(len(count), count[-1] - count[-2]))
	# # print(xy_count[len(count) - 1][len(xy_count[-2]):len(xy_count[-1])], "2")
	# # ecc = measure.regionprops(xy_count[len(count) - 1][len(xy_count[-2]):len(xy_count[-1])])
	# # print(ecc[0].eccentricity)
	# def connected_component(path):
	# img_A = cv2.imread(path)
	# gray_A = cv2.cvtColor(img_A, cv2.COLOR_BGR2GRAY) # 转换成灰度图
	# ret, thresh_A = cv2.threshold(gray_A, 20, 255, cv2.THRESH_BINARY_INV) # 灰度图转换成二值图像
	# print(thresh_A, "thresh_A")
	# thresh_A_copy = thresh_A.copy() # 复制thresh_A到thresh_A_copy
	# thresh_B = np.zeros(thresh_A.shape, np.uint8) # thresh_B大小与A相同，像素值为0
	#
	# kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (3, 3)) # 3×3结构元
	#
	# count = [] # 为了记录连通分量中的像素个数
	# xy_count = []
	# # 循环，直到thresh_A_copy中的像素值全部为0
	# while thresh_A_copy.any():
	# Xa_copy, Ya_copy = np.where(thresh_A_copy > 0) # thresh_A_copy中值为255的像素的坐标
	# thresh_B[Xa_copy[0]][Ya_copy[0]] = 255 # 选取第一个点，并将thresh_B中对应像素值改为255
	#
	# # 连通分量算法，先对thresh_B进行膨胀，再和thresh_A执行and操作（取交集）
	# for i in range(200):
	# dilation_B = cv2.dilate(thresh_B, kernel, iterations=1)
	# thresh_B = cv2.bitwise_and(thresh_A, dilation_B)
	#
	# # 取thresh_B值为255的像素坐标，并将thresh_A_copy中对应坐标像素值变为0
	# Xb, Yb = np.where(thresh_B > 0)
	# thresh_A_copy[Xb, Yb] = 0
	# xy_count.append(np.c_[Xb, Yb])
	# # 显示连通分量及其包含像素数量
	# count.append(len(Xb))
	# if len(count) == 0:
	# print("无连通分量")
	# if len(count) == 1:
	# print("第1个连通分量为{}".format(count[0]))
	# # print(np.c_[Xb, Yb], "1")
	# print(xy_count[0], "2")
	#
	# ecc = measure.regionprops(xy_count[0])
	# print(ecc[0].eccentricity)
	# if len(count) >= 2:
	# if (count[-1] - count[-2]) < 2:
	# continue
	# print("第{}个连通分量为{}".format(len(count), count[-1] - count[-2]))
	# # print(np.c_[Xb, Yb], "1")
	# print(xy_count[len(count) - 1][len(xy_count[-2]):len(xy_count[-1])], "2")
	# ecc = measure.regionprops(xy_count[len(count) - 1][len(xy_count[-2]):len(xy_count[-1])])
	# print(ecc[0].eccentricity)
	def master_control(image):
	# image = cv2.resize(image, (int(image.shape[1]*0.3), int(image.shape[0]*0.3)), interpolation=cv2.INTER_CUBIC) # 图片分辨率很大是要变小
	b, g, r = cv2.split(image) # image
	sk_frangi_img = frangi(g, scale_range=(0, 1), scale_step=0.01, beta1=1.5, beta2=0.01) # 线宽范围，步长，连接程度（越大连接越多），减少程度(越大减得越多)0.015
	sk_frangi_img = morphology.closing(sk_frangi_img, morphology.disk(1))
	sk_gabor_img_1, sk_gabor_1 = gabor(g, frequency=0.35, theta=0)
	sk_gabor_img_2, sk_gabor_2 = gabor(g, frequency=0.35, theta=45) # 越小越明显
	sk_gabor_img_3, sk_gabor_3 = gabor(g, frequency=0.35, theta=90)
	sk_gabor_img_4, sk_gabor_4 = gabor(g, frequency=0.35, theta=360) # 横向皱纹
	sk_gabor_img_1 = morphology.opening(sk_gabor_img_1, morphology.disk(2))
	sk_gabor_img_2 = morphology.opening(sk_gabor_img_2, morphology.disk(1))
	sk_gabor_img_3 = morphology.opening(sk_gabor_img_3, morphology.disk(2))
	sk_gabor_img_4 = morphology.opening(sk_gabor_img_4, morphology.disk(2))
	all_img = cv2.add(0.1 * sk_gabor_img_2, 0.9 * sk_frangi_img) # + 0.02 * sk_gabor_img_1 + 0.02 * sk_gabor_img_2 + 0.02 * sk_gabor_img_3
	all_img = morphology.closing(all_img, morphology.disk(1))
	_, all_img = cv2.threshold(all_img, 0.3, 1, 0)
	img1 = all_img
	# print(all_img, all_img.shape, type(all_img))
	# contours, image_cont = cv2.findContours(all_img, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
	# all_img = all_img + image_cont
	bool_img = all_img.astype(bool)
	label_image = measure.label(bool_img)
	count = 0
	for region in measure.regionprops(label_image):
	if region.area < 10: # or region.area > 700
	x = region.coords
	for i in range(len(x)):
	all_img[x[i][0]][x[i][1]] = 0
	continue
	if region.eccentricity > 0.98:
	count += 1
	else:
	x = region.coords
	for i in range(len(x)):
	all_img[x[i][0]][x[i][1]] = 0
	skel, distance = morphology.medial_axis(all_img.astype(int), return_distance=True)
	skels = morphology.closing(skel, morphology.disk(1))
	trans1 = skels # 细化
	return skels, count # np.uint16(skels.astype(int))
	def face_wrinkle(path):
	# result = pa.curve(path, backage)
	result = cv2.imread(path)
	img, count = master_control(result)
	print(img.astype(float))
	result[img > 0.1] = 255
	cv2.imshow("result", img.astype(float))
	cv2.waitKey(0)
	if __name__ == '__main__':
	path = r"D:\E\document\datas\line\2885.png"
	backage = r'..\Sebum\shape_predictor_81_face_landmarks.dat'
	face_wrinkle(path)

皱纹检测Wrinkle-detection相关推荐

1. 基于语义分割实现人脸图像的皱纹检测定位与分割

   前言 人脸皱纹主要区分有额纹.川字纹.眼下纹.法令纹.嘴角纹,眼角纹等,在美颜相机,智能医美等于应用领域里,需要对人脸皱纹进行检测.定位.分割,测量等. 传统图像算法的皱纹检测 1.传统算法的皱纹检测 ...

2. 行人检测(Pedestrian Detection)资源

   行人检测(Pedestrian Detection)资源 原文链接 http://hi.baidu.com/susongzhi/item/085983081b006311eafe38e7 一.论文 C ...

3. [caffe]深度学习之CNN检测object detection方法摘要介绍

   [caffe]深度学习之CNN检测object detection方法摘要介绍  2015-08-17 17:44 3276人阅读 评论(1) 收藏 举报 一两年cnn在检测这块的发展突飞猛进,下面详 ...

4. Python编程基础：第三十节 文件检测File Detection

   第三十节 文件检测File Detection 前言 实践 前言 我们通常会涉及到文件相关的操作,例如检测.读写.复制.删除等等.本节我们一起来学习文件检测相关知识,即检测指定路径下是否存在该文件. ...

5. CV之OD：计算机视觉之目标检测(Object Detection)方向的简介、使用方法、案例应用之详细攻略

   CV之OD:计算机视觉之目标检测(Object Detection)方向的简介.使用方法.案例应用之详细攻略 目录 OD目标检测的简介 1.Selective Search for Object Re ...

6. OpenCV方形检测Square Detection的实例(附完整代码)

   OpenCV方形检测Square Detection的实例 OpenCV方形检测Square Detection的实例 OpenCV方形检测Square Detection的实例 #include & ...

7. OpenCV学习(二十四 ）：角点检测(Corner Detection):cornerHarris(),goodFeatureToTrack()

   OpenCV学习(二十四 ):角点检测(Corner Detection):cornerHarris(),goodFeatureToTrack() 参考博客: Harris角点检测原理详解 Harri ...

8. 同r做一个窗口_目标检测(Object Detection)：R-CNN/SPPnet/R-FCN/Yolo/SSD

   这篇文章我是Survey目标检测(Object Detection)系列论文的一个总结. 包括R-CNN系列.SPP-net.R-FCN.YOLO系列.SSD.DenseBox等. 基本概念 目标识别 ...

9. 机器学习系列-- 异常检测(Anomaly Detection)

   目录 1.问题的动机 2.高斯分布 3.异常检测算法 4.开发和评价一个异常检测系统 5.异常检测与监督学习对比 6.特征选择 1.问题的动机 这一部分介绍异常检测(Anomaly detection ...

最新文章

1. 实现大规模图计算的算法思路
2. 为何从单体架构迁移到微服务这么难？
3. python【蓝桥杯vip练习题库】—Huffuman树
4. C#相关的DateTime操作
5. Android 系统（157）---ODM 开发用户常见需求文档（八）
6. Intelj IDEA的pom.xml显示错误can not reconnect
7. 饭团(0)：提升效率的开发工具
8. 中国剩余定理----51nod1079
9. 量化因子投资的至暗时刻，2018远比你想象得要惨烈
10. 数学建模方法之目标规划
11. 莫烦python博客_莫烦Python 4
12. java中的Properties配置文件
13. c语言从入门到秃头表情包,C语言从入门到入土表情包 - C语言从入门到入土微信表情包 - C语言从入门到入土QQ表情包 - 发表情 fabiaoqing.com...
14. 关键词词云怎么做_关键词云图生成器
15. Java获取区间随机数公式
16. python生成简单二维码_使用Python生成个性二维码
17. 防御sql注入之参数化查询
18. DTW学习（dynamic time warping）——思想、代码实现
19. 掩码认证消息（MAM）详细介绍
20. Android之极致拟物化空气质量检测仪

热门文章

1. Dijkstra算法和Floyd算法超详解以及区别
2. linux系统显卡显存容量,Linux下检查显存大小
3. Element UI学习记录之布局
4. mysql数据库查询语句_mysql数据库查询语句
5. ARM MIPS PowerPC X86 四大常见处理架构比较
6. 微信机器人接口，个人号API二次开发
7. windows版微信Hook开发SDK之C#版-微信二次开发
8. android 获得屏幕方向,Android 获取设置屏幕横竖屏
9. 柠檬班接口测试Tomcat+MySQL测试环境搭建
10. 入门使用-腾讯地图API