Influence Factors for Edge Line Estimation Accuracy Using Terrestrial Laser Scan Data

Nowadays, terrestrial laser scanning is popularly adopted in the construction industry due to its high measurement rate and accuracy. One important application of terrestrial laser scanning is to estimate the dimensions of... [ view full abstract ]

Nowadays, terrestrial laser scanning is popularly adopted in the construction industry due to its high measurement rate and accuracy. One important application of terrestrial laser scanning is to estimate the dimensions of construction components. Accurately estimating the edge lines of objects is the key to dimensional estimation. However, due to the existence of mixed pixels along the edges of objects, estimation of edge lines is always inaccurate. To tackle this problem, researchers have proposed a few models and algorithms to get rid of the influence of mixed pixels and provide more accurate estimations of edge lines. However, those research efforts did not study the factors that influence the accuracy of edge line estimation and failed to provide a guidance on how to achieve more accurate estimations. This study aims to investigate the influence factors for edge line estimation accuracy using terrestrial laser scan data. This study considers four influence factors including (1) the edge line estimation algorithm, (2) the spacing between adjacent laser spots, (3) the angle between the edge line and laser scan data, and (4) the laser spot size. The effects of the influence factors are studied based on computer simulations and implications are elaborated to provide recommendations on the selection of the influence factor values.