﻿ 基于图像内容的沥青路面病害区域分割算法
 计算机系统应用  2019, Vol. 28 Issue (2): 177-183 PDF

Asphalt Pavement Image Region Segmentation Algorithm Based on Image Content
LAN Zhang-Li, HUANG Tao, LI Zhan, KUANG Heng
School of Information Science and Engineering, Chongqing Jiaotong University, Chongqing 400074, China
Foundation item: Science and Technology Research Project of Education Commission of Chongqing Municipality (KJQN201800716)
Abstract: The resolution of the asphalt pavement disease image which collected by the CCD is too high and the area of the effective area containing information is small. A region segmentation algorithm for asphalt pavement disease image based on image content was proposed to eliminate the invalid region in the pavement image. Firstly, the original image was processed into a binary image containing the disease characteristics through a process of preprocessing and disease extraction. Then the initial traversal direction is obtained by calculating the up-to-down ratio and the left-to-right ratio of the pixels which contain information in the whole image, and counting the total number of information pixel of each row (or column). Finally, traversing from the initial traversal direction and discarding the row (or column) with the least amount of information in order to finally obtain the segmented image. In order to verify the validity and rationality of the algorithm, the image information entropy was used as the algorithm evaluation standard and compared with the traditional algorithm. The experimental results show that the proposed algorithm can keep the target information very well on the premise of effectively reducing the image resolution, and improve image information entropy.
Key words: image content     region segmentation     asphalt pavement     image processing     disease extraction

1 算法流程

 图 1 算法流程图

2 图像预处理

2.1 图像降采样

 ${Im} ag{e_{\rm{output}}}\left( {i,j} \right) = {Im} ag{e_{\rm{input}}}\left( {{2^{{k_1}}}i,{2^{{k_2}}}j} \right)$ (1)

2.2 POSHE算法

 图 2 沥青路面病害图像

POSHE算法又称为部分重叠子块直方图均衡化, 是直方图均衡化的一种改进算法, 能有效均衡图像灰度差异, 解决光照非均匀问题.

 图 3 POSHE算法过程

3 病害区域提取

OTSU又称为最大类间方差法, 是一种常用的全局阈值分割算法, 它能找出最佳分割阈值, 将图像中像素值划分为前景和背景两类. 对于校正后的图像, 该方法输出的二值图像病害区域明显, 有利于病害区域提取.

3.1 大面积虚假前景去除

 ${Im} ag{e_{\rm{output}}} = {Im} ag{e_{\rm{input}}} - {Im} ag{e_{\rm{input}}} \circ S$ (2)

3.2 小面积孤点噪声去除

 regio{n_i} = \left\{ {\begin{aligned} & 1,&{Are{a_i} \geqslant {T_a}} \\ & 0,&{Are{a_i} < {T_a}} \end{aligned}} \right. (3)

 regio{n_i} = \left\{ {\begin{aligned} & 1,&{{p_i} \leqslant {T_p}} \\ & 0,&{{p_i} > {T_p}} \end{aligned}} \right. (4)
 ${p_i} = \frac{{Are{a_i}}}{{RectangleAre{a_i}}}$ (5)

 regio{n_i} = \left\{\begin{aligned} & 1,\quad 0 < {q_i} \leqslant 1 - {T_{q1}}\quad {\rm{or}}\quad {1 + {T_{q2}} \leqslant {q_i}} \\ & 0,\quad {1 - {T_{q1}} < {q_i} < 1 + {T_{q2}}} \end{aligned} \right. (6)

4 基于图像内容的区域分割算法 4.1 算法描述

 f(i,j) = \left\{ \begin{aligned} &1, \quad {\text{含有信息}}\quad {i = 1,2, \cdots ,M}\\ & 0, \quad {\text{不含信息}}\quad {j = 1,2, \cdots ,N} \end{aligned}\right. (7)

 图 4 基于图像内容的区域分割算法流程

 \left\{ {\begin{aligned} & {{\theta _{{\text{row}}}} = N - V} \\ & {{\theta _{\rm{col}}} = M - U} \end{aligned}} \right. (8)

 ${F_j} = \sum\limits_{i = 1,2, \cdots ,M} {f(i,j)} \begin{array}{*{20}{c}} {}&{j = 1,2, \cdots ,N} \end{array}$ (9)

 ${F_i} = \sum\limits_{j = 1,2, \cdots ,N} {f(i,j)} \begin{array}{*{20}{c}} {}&{i = 1,2, \cdots ,M} \end{array}$ (10)

 \alpha = \left\{ {\begin{aligned} & {\frac{{\displaystyle\sum\limits_{i = 1, \cdots ,M\atop j = 1, \cdots ,N/2} {f\left( {i,j} \right)} }}{{\displaystyle\sum\limits_{i = 1, \cdots ,M\atop j = \left( {N + 2} \right)/2, \cdots ,N} {f\left( {i,j} \right)} }}}\quad {{{N}}{\text{为偶数}}}\\ & {\frac{{\displaystyle\sum\limits_{i = 1, \cdots ,M\atop j = 1, \cdots ,\left( {N-1} \right)/2} {f\left( {i,j} \right)} }}{{\displaystyle\sum\limits_{i = 1, \cdots ,M\atop j = \left( {N + 1} \right)/2, \cdots ,N} {f\left( {i,j} \right)} }}}\quad {{{N}}{\text{为奇数}}} \end{aligned}} \right. (11)
 \beta = \left\{ {\begin{aligned} & {\frac{{\displaystyle\sum\limits_{i = 1, \cdots ,M/2\atop j = 1, \cdots ,N} {f\left( {i,j} \right)} }}{{\displaystyle\sum\limits_{i = \left( {M + 2} \right)/2, \cdots ,M\atop j = 1, \cdots ,N} {f\left( {i,j} \right)} }}}\quad{{{M}}{\text{为偶数}}}\\ & {\frac{{\displaystyle\sum\limits_{i = 1, \cdots ,\left( {M - 1} \right)/2\atop j = 1, \cdots ,N} {f\left( {i,j} \right)} }}{{\displaystyle\sum\limits_{i = \left( {M + 1} \right)/2, \cdots ,M\atop j = 1, \cdots ,N} {f\left( {i,j} \right)} }}}\quad{{{M}}{\text{为偶数}}} \end{aligned}} \right. (12)

4.2 算法评价标准

 $E({p_0},{p_1}) = - {p_0}\log _2{{p_0}} - {p_1}\log _2{{p_1}}$ (13)
5 实验结果与分析

 图 5 沥青路面图像预处理

 图 6 沥青路面病害提取

 图 7 纵向裂缝区域分割实验结果

 图 8 横向裂缝区域分割实验结果

 图 9 网状裂缝区域分割实验结果

6 结论与展望

 [1] Sun L, Kamaliardakani M, Zhang YM. Weighted neighborhood pixels segmentation method for automated detection of cracks on pavement surface images. Journal of Computing in Civil Engineering, 2016, 30(2): 04015021. DOI:10.1061/(ASCE)CP.1943-5487.0000488 [2] 高健, 茅时群, 周宇玫, 等. 一种基于映射图像子块的图像缩小加权平均算法. 中国图象图形学报, 2006, 11(10): 1460-1463. DOI:10.3969/j.issn.1006-8961.2006.10.018 [3] 常军, 吴锡生. 基于图像子块加权缩小的自适应修正算法. 计算机工程与应用, 2009, 45(7): 181-183, 186. DOI:10.3778/j.issn.1002-8331.2009.07.054 [4] 冯杰飞, 韩慧健. 基于非均匀B样条插值算法的图像放大. 计算机应用, 2010, 30(1): 82-84, 89. [5] Mishiba K, Ikehara M, Yoshitome T. Content aware image resizing with constraint of object aspect ratio preservation. IEICE Transactions on Information and Systems, 2013, E96-D(11): 2427-2436. DOI:10.1587/transinf.E96.D.2427 [6] Niu YZ, Liu F, Li XQ, et al. Image resizing via non-homogeneous warping. Multimedia Tools and Applications, 2012, 56(3): 485-508. DOI:10.1007/s11042-010-0613-0 [7] Avidan S, Shamir A. Seam carving for content-aware image resizing. ACM Transactions on Graphics, 2007, 26(3): 10. DOI:10.1145/1276377 [8] Choi J, Kim C. Sparse seam-carving for structure preserving image retargeting. Luwer Academic Publishe, 2016, 85(2): 275-283. [9] 邹盼盼, 陆平, 朱恒亮, 等. 基于主体区域保持的图像缩放算法. 图学学报, 2016, 37(2): 230-236. [10] 林晓, 沈洋, 马利庄, 等. 显著物体形状结构保持的图像缩放方法. 计算机科学, 2014, 41(12): 288-292. DOI:10.11896/j.issn.1002-137X.2014.12.062 [11] 宋蓓蓓, 韦娜. FCM分割和形态学的沥青路面图像裂缝提取. 计算机工程与应用, 2013, 49(4): 31-34. DOI:10.3778/j.issn.1002-8331.1208-0201 [12] 文立. 改进的灰度校正算法在路面裂缝图像预处理中应用. 计算机系统应用, 2015, 24(2): 220-223. DOI:10.3969/j.issn.1003-3254.2015.02.041 [13] Yang W, Xiao ZT, Wang QJ, et al. A method for improving the definition of scene in fog image. 2009 International Conference on Computational Intelligence and Software Engineering. Wuhan, China. 2009. 1–4.