Reviewing Drone Imagery for Infrastructure

In a paper published in the journal Automation in Construction, researchers conducted a tertiary study to review secondary literature on computer vision (CV) applications for infrastructure management using drone imagery. Analyzing 57 secondary studies from 2018 to 2023 across three databases, they assessed publication trends and quality using DARE criteria and identified six major application areas.

Study: Reviewing Drone Imagery for Infrastructure. Image Credit: NorCalStockMedia/Shutterstock.com
Study: Reviewing Drone Imagery for Infrastructure. Image Credit: NorCalStockMedia/Shutterstock.com

Background

Past work in tertiary studies aimed at synthesizing secondary literature on infrastructure management using drone imagery has faced challenges due to the broad scope and outdated coverage. Existing reviews, including one focused on unmanned aircraft systems from 2011 to 2019, lacked specificity and failed to capture recent advancements.

Additionally, these reviews often needed help integrating findings from diverse publication venues, making it difficult to effectively identify and address emerging trends. The rapid evolution of drone technology further complicates the task, as recent innovations may need to be adequately reflected in older reviews.

Methodology and Selection

The tertiary study utilized the systematic review methodology outlined by Kitchenham and Charters. This approach involved formulating research questions, defining inclusion and exclusion criteria, constructing and validating search strings, and selecting pertinent databases. The detailed methodology included generating search terms based on keywords from known papers, validating the search string against a quasi-gold standard, and using Scopus, the Web of Science, and ScienceDirect to ensure comprehensive coverage of recent literature from 2018 to 2023.

Research questions focused on understanding trends, patterns, and quality in secondary studies related to CV applications in drone imagery for infrastructure management. The questions explored demographic trends, publication distribution, quality patterns over time, and prevalent applications of CV methods. These questions were mapped to specific study sections for organized analysis, ensuring a clear link between research questions and corresponding sections.

The selection process involved reviewing abstracts and full-text articles based on inclusion and exclusion criteria, followed by a quality assessment using the database of abstracts of reviews of effects (DARE) criteria. Initial searches yielded 110 studies, updated in subsequent iterations to 88 unique studies. After screening and applying quality criteria, 57 primary studies were selected for final analysis.

Study Overview

The study employed Scopus, Web of Science, and ScienceDirect for a comprehensive literature search covering secondary studies from 2018 to 2023. Their publishers categorized the selected studies and assigned unique codes for easy reference.

The number of primary studies cited by each secondary study was carefully examined, with references reviewed to ensure accuracy. Citation analysis, conducted via Google Scholar and normalized by publication year, provided insights into the research impact. The yearly distribution of publications revealed significant trends, with the highest number of publications in 2022, followed by fluctuations in other years.

The review categorized the 57 secondary studies into systematic literature reviews (SLRs), comprehensive reviews, and narrative reviews. Systematic reviews showed the highest adherence to quality criteria, achieving an average database of abstracts of reviews of effects (DARE) score of 3.80. In contrast, comprehensive and narrative reviews exhibited lower average ratings of 2.60 and 2.59, respectively. It reflects varying methodological rigor across review types.

The quality of studies by publication year showed no significant correlation between publication year and quality. Overall, SLRs demonstrated superior quality, highlighting their rigorous methodology compared to comprehensive and narrative reviews.

Application Overview

To address research question 3 (RQ3), themes from secondary studies were coded using non-visual interactive vocabulary (NVIVO) software, leading to a comprehensive understanding of various infrastructural elements. The team analyzed each study's objectives and research questions and categorized them into themes through systematic coding. This analysis revealed seven significant application categories:

The most discussed topics were bridge inspection, building inspection, construction management, railways management, roads and pavements management, traffic management, and miscellaneous applications. Roads and pavement management was followed by bridge and building inspections.

Each category included studies covering specific aspects, such as road distress detection methods, bridge inspection automation, and building inspection faults. The thematic analysis helped answer RQ3.1 by identifying the prevalence and focus of different infrastructure management applications.

The review further detailed each application area, highlighting the image processing techniques. For instance, various methods like deep learning (DL) and structure-from-motion were used for distress detection in road management. Bridge inspection studies focused on automation and applying deep learning for damage detection.

Building inspections covered structural deficiencies, roofing issues, and facade evaluations, with deep learning becoming prominent. Traffic management involved vehicle detection, tracking, and flow analysis, utilizing deep learning and real-time data processing. Construction management included real-time site monitoring and 3D mapping, while railway management addressed track defects and natural hazards.

Miscellaneous applications covered diverse uses like pipeline inspections, power infrastructure, and waste management. This thematic approach provided insights into the varied applications of drone technology and image processing across different infrastructure domains.

Conclusion

To sum up, a systematic tertiary study analyzed drone imagery for infrastructure management, reviewing 57 secondary studies from 2018 to 2023. It synthesized research trends, application types, and publication details while evaluating study quality. The study also highlighted challenges and future directions and served as a central reference for researchers and practitioners in the field.

Journal reference:
Silpaja Chandrasekar

Written by

Silpaja Chandrasekar

Dr. Silpaja Chandrasekar has a Ph.D. in Computer Science from Anna University, Chennai. Her research expertise lies in analyzing traffic parameters under challenging environmental conditions. Additionally, she has gained valuable exposure to diverse research areas, such as detection, tracking, classification, medical image analysis, cancer cell detection, chemistry, and Hamiltonian walks.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Chandrasekar, Silpaja. (2024, August 07). Reviewing Drone Imagery for Infrastructure. AZoAi. Retrieved on October 08, 2024 from https://www.azoai.com/news/20240807/Reviewing-Drone-Imagery-for-Infrastructure.aspx.

  • MLA

    Chandrasekar, Silpaja. "Reviewing Drone Imagery for Infrastructure". AZoAi. 08 October 2024. <https://www.azoai.com/news/20240807/Reviewing-Drone-Imagery-for-Infrastructure.aspx>.

  • Chicago

    Chandrasekar, Silpaja. "Reviewing Drone Imagery for Infrastructure". AZoAi. https://www.azoai.com/news/20240807/Reviewing-Drone-Imagery-for-Infrastructure.aspx. (accessed October 08, 2024).

  • Harvard

    Chandrasekar, Silpaja. 2024. Reviewing Drone Imagery for Infrastructure. AZoAi, viewed 08 October 2024, https://www.azoai.com/news/20240807/Reviewing-Drone-Imagery-for-Infrastructure.aspx.

Comments

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of AZoAi.
Post a new comment
Post

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
Deep Learning Transforms Solar Cell Design