Sensor location method for surface inspection and defect measurement

This work presents a geometric method to analyze the influence of laser triangulation sensor positioning on the accuracy of surface defect measurements. The novelty of the approach lies in a parametric analytical model that quantifies measurement error as a function of defect geometry (width and dep...

Descripción completa

Detalles Bibliográficos
Autores: Calle Herrero, Francisco Javier de la|||0000-0002-5546-2005, González Lema, Darío|||0000-0003-0981-7925, Usamentiaga Fernández, Rubén|||0000-0003-0551-3203
Tipo de recurso: artículo
Fecha de publicación:2026
País:España
Institución:Universidad de Oviedo (UNIOVI)
Repositorio:RUO. Repositorio Institucional de la Universidad de Oviedo
Idioma:inglés
OAI Identifier:oai:dnet:ruo_________::33558c873f71d312a2806cfb29dc19ee
Acceso en línea:https://hdl.handle.net/10651/83276
https://dx.doi.org/10.1016/J.MEASUREMENT.2025.119183
Access Level:acceso abierto
Palabra clave:Computer vision
Defect detection
Laser sensor
Measurement
Surface inspection
Descripción
Sumario:This work presents a geometric method to analyze the influence of laser triangulation sensor positioning on the accuracy of surface defect measurements. The novelty of the approach lies in a parametric analytical model that quantifies measurement error as a function of defect geometry (width and depth), the working distance, and the acquisition angle. The method predicts an analytical error, enabling the estimation of measurement deviations before experimentation. Validation was performed using 3D-printed and aluminum samples with controlled defects. The analytical predictions agree with experimental measurements from a Gocator 2350 sensor, with average discrepancies of less than 0.1 mm in depth and width. A second validation on a real rail, representative of long products with strict surface quality standards, confirmed the robustness of the approach in industrial conditions. Across multiple acquisition angles and defect orientations, the analytical model consistently reproduced the empirical trends within the sensor's precision. These results demonstrate the method's reliability and practical relevance for optimizing sensor placement in industrial surface inspection systems.