【youcans 的 OpenCV 例程200篇】127. 形态算法之细化

欢迎关注 『youcans 的 OpenCV 例程 200 篇』 系列,持续更新中
欢迎关注 『youcans 的 OpenCV学习课』 系列,持续更新中


【youcans 的 OpenCV 例程 200 篇】127. 形态算法之细化

3. 形态学算法

形态学处理的主要应用是提取图像中用来表示和描述形状的元素和成分,例如提取边界、连通分量、凸壳和区域骨架。


3.5 细化

细化是将图像的线条从多像素宽度减少到单位像素宽度的过程,也称为"骨架化"、“中轴转换"和"对称轴转换”。

形态骨架(morphological skeleton)是一种细化的结构,指图像的骨骼部分,用于描述物体的几何形状和拓扑结构,是目标物体重要的拓扑描述。图像的细化是对二值图像进行骨架提取,删除不需要的轮廓点,只保留其骨架点。

数学形态学细化算法的主要思想是源于击中击不中变换的形态学算法。在给定一系列具有一定形状的结构元素后,顺序循环地删除满足击中变换的像素。结构元 B 对前景像素集合 A 的细化,可以根据击中-击不中变换来定义:
A ⊗ B = A − ( A ⊛ B ) = A ∩ ( A ⊛ B ) c { B } = { B 1 , B 2 , . . . B n } A ⊗ { B } = ( ( . . . ( ( A ⊗ B 1 ) ⊗ B 2 ) . . . ) ⊗ B n ) A \otimes B = A - (A \circledast B) = A \cap (A \circledast B)^c \\ \{ B \} = \{ B^1, B^2, ...B^n\} \\ A \otimes \{ B \} = ((...((A \otimes B^1)\otimes B^2)...)\otimes B^n) AB=A(AB)=A(AB)c{B}={B1,B2,...Bn}A{B}=((...((AB1)B2)...)Bn)

细化过程就是对图像不断重复地逐层边界像素的过程,目标物体随着细化的进行有规律地缩小,但是目标图像边界线的连接性,方向性和特征点不变,最终使变换图像成为单像素宽的图像骨架。

细化算法应满足以下条件:
(1)将条形区域变成一条细线;
(2)细线应尽可能位于原条形区域的中间;
(3)细化应保持原图像的拓扑特性。

典型的细化算法有 Hilditch 细化算法、Deutch 细化算法、Pavlidis 细化算法、 Zhang 快速并行算法等。

细化可以突出目标的形状特点和拓扑结构,减少冗余的数据和信息,留下足够的有用信息来进行拓扑分析、形状分析或者原始对象的还原。图像细化广泛应用于文字识别、零件形状识别、指纹分类、印刷电路板检测、染色体分析等领域。


例程 10.15:形态算法之细化算法

细化是从原来的图中去掉一些点,但仍然保持原来的形状。判断一个点能否被去掉是基于 8邻域的取值进行判断:内部点不能删除,孤立点不能删除,直线端点不能删除,边界点如果去掉后不增加连通分量则可以删除。

    # 10.15 形态算法之细化算法
    def thinning(image):
        array = [0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, \
                 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, \
                 0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, \
                 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, \
                 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, \
                 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, \
                 1, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, \
                 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, \
                 0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, \
                 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, \
                 0, 0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, \
                 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, \
                 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, \
                 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, \
                 1, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, \
                 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0, 0]

        h, w = image.shape[0], image.shape[1]
        imgThin = image.copy()
        for i in range(h):
            for j in range(w):
                if image[i, j] == 0:
                    a = np.ones((9,), dtype=np.int)
                    for k in range(3):
                        for l in range(3):
                            if -1<(i-1+k)<h and -1<(j-1+l)<w and imgThin[i-1+k,j-1+l]==0:
                                a[k*3+l] = 0
                    sum = a[0]*1 + a[1]*2 + a[2]*4 + a[3]*8 + a[5]*16 + a[6]*32 + a[7]*64 + a[8]*128
                    imgThin[i, j] = array[sum] * 255
        return imgThin

    # 图像为灰度图像,背景为白色(255),被细化物体为黑色(0)
    image = cv2.imread("../images/imgNetrope.png", flags=0)  # flags=0 灰度图像
    ret, binary = cv2.threshold(image, 205, 255, cv2.THRESH_BINARY)  # 二值化处理

    imgThin = thinning(binary)  # 细化算法

    plt.figure(figsize=(9, 6))
    plt.subplot(131), plt.axis('off'), plt.title("origin")
    plt.imshow(image, cmap='gray', vmin=0, vmax=255)
    plt.subplot(132), plt.title("binary"), plt.axis('off')
    plt.imshow(binary, cmap='gray', vmin=0, vmax=255)
    plt.subplot(133), plt.title("thinnning"), plt.axis('off')
    plt.imshow(imgThin, cmap='gray', vmin=0, vmax=255)
    plt.tight_layout()
    plt.show()

在这里插入图片描述


(本节完)


版权声明:

youcans@xupt 原创作品,转载必须标注原文链接:(https://blog.csdn.net/youcans/article/details/123457160)

Copyright 2022 youcans, XUPT
Crated:2022-3-12


欢迎关注 『youcans 的 OpenCV 例程 200 篇』 系列,持续更新中
欢迎关注 『youcans 的 OpenCV学习课』 系列,持续更新中

【youcans 的 OpenCV 例程200篇】01. 图像的读取(cv2.imread)
【youcans 的 OpenCV 例程200篇】02. 图像的保存(cv2.imwrite)
【youcans 的 OpenCV 例程200篇】03. 图像的显示(cv2.imshow)
【youcans 的 OpenCV 例程200篇】04. 用 matplotlib 显示图像(plt.imshow)
【youcans 的 OpenCV 例程200篇】05. 图像的属性(np.shape)
【youcans 的 OpenCV 例程200篇】06. 像素的编辑(img.itemset)
【youcans 的 OpenCV 例程200篇】07. 图像的创建(np.zeros)
【youcans 的 OpenCV 例程200篇】08. 图像的复制(np.copy)
【youcans 的 OpenCV 例程200篇】09. 图像的裁剪(cv2.selectROI)
【youcans 的 OpenCV 例程200篇】10. 图像的拼接(np.hstack)
【youcans 的 OpenCV 例程200篇】11. 图像通道的拆分(cv2.split)
【youcans 的 OpenCV 例程200篇】12. 图像通道的合并(cv2.merge)
【youcans 的 OpenCV 例程200篇】13. 图像的加法运算(cv2.add)
【youcans 的 OpenCV 例程200篇】14. 图像与标量相加(cv2.add)
【youcans 的 OpenCV 例程200篇】15. 图像的加权加法(cv2.addWeight)
【youcans 的 OpenCV 例程200篇】16. 不同尺寸的图像加法
【youcans 的 OpenCV 例程200篇】17. 两张图像的渐变切换
【youcans 的 OpenCV 例程200篇】18. 图像的掩模加法
【youcans 的 OpenCV 例程200篇】19. 图像的圆形遮罩
【youcans 的 OpenCV 例程200篇】20. 图像的按位运算
【youcans 的 OpenCV 例程200篇】21. 图像的叠加
【youcans 的 OpenCV 例程200篇】22. 图像添加非中文文字
【youcans 的 OpenCV 例程200篇】23. 图像添加中文文字
【youcans 的 OpenCV 例程200篇】24. 图像的仿射变换
【youcans 的 OpenCV 例程200篇】25. 图像的平移
【youcans 的 OpenCV 例程200篇】26. 图像的旋转(以原点为中心)
【youcans 的 OpenCV 例程200篇】27. 图像的旋转(以任意点为中心)
【youcans 的 OpenCV 例程200篇】28. 图像的旋转(直角旋转)
【youcans 的 OpenCV 例程200篇】29. 图像的翻转(cv2.flip)
【youcans 的 OpenCV 例程200篇】30. 图像的缩放(cv2.resize)
【youcans 的 OpenCV 例程200篇】31. 图像金字塔(cv2.pyrDown)
【youcans 的 OpenCV 例程200篇】32. 图像的扭变(错切)
【youcans 的 OpenCV 例程200篇】33. 图像的复合变换
【youcans 的 OpenCV 例程200篇】34. 图像的投影变换
【youcans 的 OpenCV 例程200篇】35. 图像的投影变换(边界填充)
【youcans 的 OpenCV 例程200篇】36. 直角坐标与极坐标的转换
【youcans 的 OpenCV 例程200篇】37. 图像的灰度化处理和二值化处理
【youcans 的 OpenCV 例程200篇】38. 图像的反色变换(图像反转)
【youcans 的 OpenCV 例程200篇】39. 图像灰度的线性变换
【youcans 的 OpenCV 例程200篇】40. 图像分段线性灰度变换
【youcans 的 OpenCV 例程200篇】41. 图像的灰度变换(灰度级分层)
【youcans 的 OpenCV 例程200篇】42. 图像的灰度变换(比特平面分层)
【youcans 的 OpenCV 例程200篇】43. 图像的灰度变换(对数变换)
【youcans 的 OpenCV 例程200篇】44. 图像的灰度变换(伽马变换)
【youcans 的 OpenCV 例程200篇】45. 图像的灰度直方图
【youcans 的 OpenCV 例程200篇】46. 直方图均衡化
【youcans 的 OpenCV 例程200篇】47. 图像增强—直方图匹配
【youcans 的 OpenCV 例程200篇】48. 图像增强—彩色直方图匹配
【youcans 的 OpenCV 例程200篇】49. 图像增强—局部直方图处理
【youcans 的 OpenCV 例程200篇】50. 图像增强—直方图统计量图像增强
【youcans 的 OpenCV 例程200篇】51. 图像增强—直方图反向追踪
【youcans 的 OpenCV 例程200篇】52. 图像的相关与卷积运算
【youcans 的 OpenCV 例程200篇】53. Scipy 实现图像二维卷积
【youcans 的 OpenCV 例程200篇】54. OpenCV 实现图像二维卷积
【youcans 的 OpenCV 例程200篇】55. 可分离卷积核
【youcans 的 OpenCV 例程200篇】56. 低通盒式滤波器
【youcans 的 OpenCV 例程200篇】57. 低通高斯滤波器
【youcans 的 OpenCV 例程200篇】58. 非线性滤波—中值滤波
【youcans 的 OpenCV 例程200篇】59. 非线性滤波—双边滤波
【youcans 的 OpenCV 例程200篇】60. 非线性滤波—联合双边滤波
【youcans 的 OpenCV 例程200篇】61. 导向滤波(Guided filter)
【youcans 的 OpenCV 例程200篇】62. 图像锐化——钝化掩蔽
【youcans 的 OpenCV 例程200篇】63. 图像锐化——Laplacian 算子
【youcans 的 OpenCV 例程200篇】64. 图像锐化——Sobel 算子
【youcans 的 OpenCV 例程200篇】65. 图像锐化——Scharr 算子
【youcans 的 OpenCV 例程200篇】66. 图像滤波之低通/高通/带阻/带通
【youcans 的 OpenCV 例程200篇】67. 空间域图像增强的综合应用
【youcans 的 OpenCV 例程200篇】68. 空间域图像增强的综合应用
【youcans 的 OpenCV 例程200篇】69. 连续非周期信号的傅立叶系数
【youcans 的 OpenCV 例程200篇】70. 一维连续函数的傅里叶变换
【youcans 的 OpenCV 例程200篇】71. 连续函数的取样
【youcans 的 OpenCV 例程200篇】72. 一维离散傅里叶变换
【youcans 的 OpenCV 例程200篇】73. 二维连续傅里叶变换
【youcans 的 OpenCV 例程200篇】74. 图像的抗混叠
【youcans 的 OpenCV 例程200篇】75. Numpy 实现图像傅里叶变换
【youcans 的 OpenCV 例程200篇】76. OpenCV 实现图像傅里叶变换
【youcans 的 OpenCV 例程200篇】77. OpenCV 实现快速傅里叶变换
【youcans 的 OpenCV 例程200篇】78. 频率域图像滤波基础
【youcans 的 OpenCV 例程200篇】79. 频率域图像滤波的基本步骤
【youcans 的 OpenCV 例程200篇】80. 频率域图像滤波详细步骤
【youcans 的 OpenCV 例程200篇】81. 频率域高斯低通滤波器
【youcans 的 OpenCV 例程200篇】82. 频率域巴特沃斯低通滤波器
【youcans 的 OpenCV 例程200篇】83. 频率域低通滤波:印刷文本字符修复
【youcans 的 OpenCV 例程200篇】84. 由低通滤波器得到高通滤波器
【youcans 的 OpenCV 例程200篇】85. 频率域高通滤波器的应用
【youcans 的 OpenCV 例程200篇】86. 频率域滤波应用:指纹图像处理
【youcans 的 OpenCV 例程200篇】87. 频率域钝化掩蔽
【youcans 的 OpenCV 例程200篇】88. 频率域拉普拉斯高通滤波
【youcans 的 OpenCV 例程200篇】89. 带阻滤波器的传递函数
【youcans 的 OpenCV 例程200篇】90. 频率域陷波滤波器
【youcans 的 OpenCV 例程200篇】91. 高斯噪声、瑞利噪声、爱尔兰噪声
【youcans 的 OpenCV 例程200篇】92. 指数噪声、均匀噪声、椒盐噪声
【youcans 的 OpenCV 例程200篇】93. 噪声模型的直方图
【youcans 的 OpenCV 例程200篇】94. 算术平均滤波器
【youcans 的 OpenCV 例程200篇】95. 几何均值滤波器
【youcans 的 OpenCV 例程200篇】96. 谐波平均滤波器
【youcans 的 OpenCV 例程200篇】97. 反谐波平均滤波器
【youcans 的 OpenCV 例程200篇】98. 统计排序滤波器
【youcans 的 OpenCV 例程200篇】99. 修正阿尔法均值滤波器
【youcans 的 OpenCV 例程200篇】100. 自适应局部降噪滤波器
【youcans 的 OpenCV 例程200篇】101. 自适应中值滤波器
【youcans 的 OpenCV 例程200篇】102. 陷波带阻滤波器的传递函数
【youcans 的 OpenCV 例程200篇】103. 陷波带阻滤波器消除周期噪声干扰
【youcans 的 OpenCV 例程200篇】104. 运动模糊退化模型
【youcans 的 OpenCV 例程200篇】105. 湍流模糊退化模型
【youcans 的 OpenCV 例程200篇】106. 退化图像的逆滤波
【youcans 的 OpenCV 例程200篇】107. 退化图像的维纳滤波
【youcans 的 OpenCV 例程200篇】108. 约束最小二乘方滤波
【youcans 的 OpenCV 例程200篇】109. 几何均值滤波
【youcans 的 OpenCV 例程200篇】110. 投影和雷登变换
【youcans 的 OpenCV 例程200篇】111. 雷登变换反投影重建图像
【youcans 的 OpenCV 例程200篇】112. 滤波反投影重建图像
【youcans 的 OpenCV 例程200篇】113. 形态学操作之腐蚀
【youcans 的 OpenCV 例程200篇】114. 形态学操作之膨胀
【youcans 的 OpenCV 例程200篇】115. 形态学操作之开运算
【youcans 的 OpenCV 例程200篇】116. 形态学操作之闭运算
【youcans 的 OpenCV 例程200篇】117. 形态学操作之顶帽运算
【youcans 的 OpenCV 例程200篇】118. 形态学操作之底帽运算
【youcans 的 OpenCV 例程200篇】119. 图像的形态学梯度
【youcans 的 OpenCV 例程200篇】120. 击中-击不中变换
【youcans 的 OpenCV 例程200篇】121. 击中-击不中用于特征识别
【youcans 的 OpenCV 例程200篇】122. 形态算法之边界提取
【youcans 的 OpenCV 例程200篇】123. 形态算法之孔洞填充
【youcans 的 OpenCV 例程200篇】124. 孔洞填充的泛洪算法
【youcans 的 OpenCV 例程200篇】125. 形态算法之提取连通分量
【youcans 的 OpenCV 例程200篇】126. 形态算法之凸壳
【youcans 的 OpenCV 例程200篇】127. 形态算法之细化

  • 0
    点赞
  • 9
    收藏
    觉得还不错? 一键收藏
  • 打赏
    打赏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

youcans_

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值