A Combinatory Approach with Morphological Component Analysis to Achieve a Submillimeter Width Crack Detection Designed for Early Prediction of the Concrete Degradation in the Civil Infrastructure

Dixit Ankur,

WEKO3

lat lon distance

[[sub_check.contents]]

[[sub_radio.contents]]

Field does not validate

[[sub_attr.contents]]　

インデックスツリー

アイテム

コンクリート構造物劣化早期予測のためのミリ以下幅ひび割れ検知を可能にするMCA複合解析に関する研究

https://doi.org/10.18997/0002000925

名前 / ファイル	ライセンス	アクション
sei_o_10.pdf (31.1 MB)

アイテムタイプ

学位論文 = Thesis or Dissertation(1)

公開日

2024-08-27

資源タイプ

資源タイプ識別子

http://purl.org/coar/resource_type/c_db06

資源タイプ

doctoral thesis

タイトル

A Combinatory Approach with Morphological Component Analysis to Achieve a Submillimeter Width Crack Detection Designed for Early Prediction of the Concrete Degradation in the Civil Infrastructure

言語

タイトル

コンクリート構造物劣化早期予測のためのミリ以下幅ひび割れ検知を可能にするMCA複合解析に関する研究

言語

eng

著者

Dixit Ankur,

抄録

内容記述タイプ

Abstract

内容記述

Detection of cracks on the surface of concrete construction is critical for the maintenance of social infrastructure and requires research efforts in civil engineering and computer vision. For sake of a significant reduction in the workloads of expert human inspectors, small size crack detections such as cracks of less than two millimeters in width is highly important for early preventions of severe accidents; however, an accurate automatic detection in the level remains unsolved with conventional methods.
Traditional schemes like PCA have relied on linear decomposition for the separation of target signal and noise components. Recent advancement in signal decomposition focuses on the enhancement of linearity in the separation by introducing a set of nonlinear basis functions to represent the raw signal even when multiple factors are mixed in a nonlinear manner. In this sense, MCA is a core technique to be able to isolate target components to represent submillimeter-width cracks from others.
In this dissertation work, we hypothesized a vision based automatic concrete crack detection method that uses morphological component analysis (MCA) technique to detect the fine cracks even for two millimeter width. In this method, traditional MCA technique is modified to decompose the image into its coarse and fine components with selected dictionaries and appropriated pre- and post-processes.
This modified MCA technique is combined with modified anisotropic diffusion and Sobel edge detector to develop the crack detection algorithm. First of all, the main focus of this dissertation work is the selection of the best dictionaries in MCA to find the coarse and fine components of an image, which is an important parameter to enhance the performance of crack detection algorithm.
For this selection, the results of crack detection algorithm in the presence of multiple combinations of dictionaries in MCA are compared with cracks drawn manually on the images. Hit rates and coverage ratios for a set of large number of images are calculated to measure the accuracy and to validate the performance of crack detection algorithm in the presence of various dictionaries in MCA.
The results of comparisons show that by using stationary wavelet transform and discrete wavelet transform as the dictionaries in MCA, our algorithm is able to detect cracks efficiently in the image. Second target of this dissertation work is to develop the crack detection algorithm which is able to find the minor (crack width ≤ 2 mm or 10 pixels), medium (10 pixels < crack width ≤ 20 pixels) and large size cracks (crack width > 20 pixels) present in the image. Our model which is the combination of modified versions of MCA and anisotropic diffusion with Sobel edge detector, used to find the cracks in the images.
Based on hit rages and coverage ratios analysis, the proposed method is compared with traditional crack detection methods named as anisotropic diffusion and image morphology. The cracks detection results, hit rates and coverage ratios show that the proposed method is efficient to detect cracks efficiently on the bridge surface with high accuracy compared to other methods.

内容記述タイプ

TableOfContents

内容記述

1 Introduction| 2 Introduction to Sparse coding and Morphological Component Analysis(MCA)| 3 A Concrete Crack Inspection Embedded in the Drone-Based System by Using Sub-Pixel Width Estimation and Morphological Component Analysis| 4 An Image Processing Mechanism for Aerial Inspection Robots to Detect Submillimeter-Width Concrete Cracks in Social Infrastructures| 5 A Validation Analysis of the Submillimeter-Width Concrete Crack Detection Based on Morphological Component Analysis and Anisotropic Diffusion| 6 Summary

備考

内容記述タイプ

Other

内容記述

九州工業大学博士学位論文学位記番号:生工博乙第10号学位授与年月日:令和6年6月28日

学位授与番号

乙第10号

学位名

博士（工学）

学位授与年月日

2024-06-28

学位授与機関

学位授与機関識別子Scheme

kakenhi

学位授与機関識別子

17104

学位授与機関名

九州工業大学

言語

学位授与年度

内容記述タイプ

Other

内容記述

令和6年度

出版タイプ

VoR

出版タイプResource

http://purl.org/coar/version/c_970fb48d4fbd8a85

アクセス権

open access

アクセス権URI

http://purl.org/coar/access_right/c_abf2

ID登録

10.18997/0002000925

ID登録タイプ

JaLC

戻る

views

See details

	Views

Versions

Ver.1

2024-08-27 01:41:45.846584

Show All versions

Cite as

Other

エクスポート

OAI-PMH

JPCOAR 2.0
JPCOAR 1.0
DublinCore
DDI

Other Formats

インデックスリンク

インデックスツリー

アイテム

コンクリート構造物劣化早期予測のためのミリ以下幅ひび割れ検知を可能にするMCA複合解析に関する研究

× Dixit Ankur,

Versions

Share

Cite as

Other

エクスポート

コミュニティ

メニューを最小化

インデックスリンク

インデックスツリー

アイテム

コンクリート構造物劣化早期予測のためのミリ以下幅ひび割れ検知を可能にするMCA複合解析に関する研究

× Dixit Ankur,

Versions

Share

Cite as

Other

エクスポート

コミュニティ