Citation

Azmawi, Azib Jazman (2025) Automated assessment of fibre box installation using deep learning and geometric image analysis. Masters thesis, Multimedia University.

Full text not available from this repository.

Abstract

The telecommunication industry in Malaysia plays a pivotal role in today’s digital age by providing essential Internet and communication services. Among various installation available, Internet Service Providers (ISPs) are increasingly adopting on Fibre Distribution Panels (FDPs) to deliver high-speed Internet to residential and enterprise users. For each new installation, an assessment is required to ensure safety and readiness of these setups by following proper business practices to maintain product reliability and user satisfaction. A key aspect of this assessment is verifying that the FDP follows a minimum spacing around its component. However, current assessment method is inefficient and resource-intensive as it relies on manual measurements and visual inspection, which are prone to inconsistency and human error. While automated safety assessment systems have been widely implemented in various industries, their application in the telecommunication sector remains largely unexplored. Thus, this study addresses this limitation by leveraging computer vision techniques to automate the safety distance calculation for FDP installations. The proposed method uses computer vision and deep learning techniques to automatically measure distance between two objects and evaluate the overall installation quality. Experimental results using real-world data demonstrate that the proposed method achieves 60% accuracy in component width retrieval and yields 80% agreement with human expert answers on violated installations. Additionally, our method significantly reduces inspection time, making the assessment process 90% faster as compared to traditional method. By automating FDP installation assessments, this research contributes to enhancing the efficiency, safety and service quality within the telecommunication industry, particularly in FDP evaluations, ensuring its customers receive reliable Internet access in a safe and timely manner.

Item Type:	Thesis (Masters)
Additional Information:	Call No.: Q325.73 .A95 2025
Uncontrolled Keywords:	Deep learning (Machine learning)
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	22 May 2026 05:24
Last Modified:	22 May 2026 05:24
URII:	http://shdl.mmu.edu.my/id/eprint/15901

Downloads

Edit (login required)