Abstract
An approach is explored for modernizing the determination of bulk density of compacted asphalt specimens. It is based on calculating the bulk volume of the specimen in a three-dimensional model reconstructed from its images. The paper presents the basics of image-based modeling, founded upon the science of photogrammetry and computer vision. Next, a demonstrative application is described, in which a field core is photographed from many viewpoints with a
consumer grade camera, and the images are combined into a sparse point cloud. This cloud is subsequently ‘meshed’ with planar polygons into a closed 3D shape and its volume calculated. It was found that the model-core volume was very close to that measured with a traditional liquid displacement approach. It was also found that while the volume was relatively insensitive to the quantity and quality of the images used for the reconstruction, the computational time varied significantly from minutes to hours. Based on the favorable findings of this limited application, the approach is deemed promising and viable, worthy of more in-depth examination.
consumer grade camera, and the images are combined into a sparse point cloud. This cloud is subsequently ‘meshed’ with planar polygons into a closed 3D shape and its volume calculated. It was found that the model-core volume was very close to that measured with a traditional liquid displacement approach. It was also found that while the volume was relatively insensitive to the quantity and quality of the images used for the reconstruction, the computational time varied significantly from minutes to hours. Based on the favorable findings of this limited application, the approach is deemed promising and viable, worthy of more in-depth examination.
Original language | English |
---|---|
Journal | International Journal of Transportation Science and Technology |
Volume | 2 |
Issue number | 1 |
Pages (from-to) | 1-14 |
ISSN | 2046-0430 |
Publication status | Published - 2013 |
Externally published | Yes |