This thesis tackles the crucial need for automated height measurement systems in the automobile manufacturing industry, with a particular emphasis on the industrialization of a computer vision-based solution at the Volvo Trucks Plant in Tuve, Gothenburg. Traditional manual measurement methods, which use spirit levels and measuring poles, suffer from operator variability, long processing times (about 100 seconds per vehicle), and limited documentation capabilities, reducing efficiency and quality control in manufacturing operations.

Using advanced computer vision and machine learning techniques, the study converts a proof-of-concept prototype into a production-ready system. The designed system accurately delineates truck borders across different production of some models, with measuring precision within ±5 cm tolerance. This is performed via a sophisticated technique that combines high-resolution photography, deep learning-driven segmentation, and strong reference calibration with fiducial markers to ensure accurate pixel-to-physical measurement conversion.

Key advances include the development of a full industrial implementation frame-work that includes hardware requirements, software architecture, environmental controls, and user interfaces while taking into account production restrictions and longterm supportability. The system’s server architecture allows for consistent camera data gathering, efficient image processing, safe result storage, and seamless connection with existing factory information systems.

Extensive testing on a variety of truck models, combinations, and climatic circumstances verifies the system’s measuring accuracy, repeatability, and reliability.

The installation considerably saves measurement time while eliminating human error and ensuring consistent documentation for trend analysis. Beyond its immediate use, this study provides valuable insights into industrial automation and computer vision applications, with implications for logistics, intelligent transportation systems, and quality control procedures in the car manufacturing sector.

This work demonstrates how targeted technology innovation can address specific industrial difficulties while also supporting broader industry trends such as automation, quality control, and worldwide regulatory compliance in the context of Industry 4.0.