Pipe Measurement: Photogrammetry, Recognition, and Reconstruction

Industrial piping is designed and constructed to transport fluids from one location to another and also connects industrial equipment. However, with increasingly strict safety requirements in pipe construction and maintenance, problems exposed in the traditional piping construction such as the lack of comprehensive overall planning, high safety risk, lower construction efficiency, and the degeneration of pipes. This, therefore, requires owners, managers, and designers to define suitable maintenance strategies and modernize the piping to ensure that infrastructure remains serviceable and to prevent potential safety hazards while also extending the piping system's lifespan.

Automatic Feature Selection in Digital Photogrammetry

Current point cloud construction methods are not suited to complex piping systems, and many point cloud merging methods perform poorly in complex piping environments. To provide critical functions for constructing pipe point clouds from digital photogrammetry, an automatic feature selection network (FSNet) is proposed. First, a digital photogrammetry method combining FSNet with handcrafted features is proposed according to the characteristics of the piping environment. Second, a piping dataset of various piping scenes is constructed using the developed data capturing device. A training dataset and the corresponding network output categories are determined according to pose estimation performance using different image features. Finally, experiments conducted on the final test dataset indicate that, together, digital photogrammetry and FSNet can improve accuracy, flexibility, and processing speed.

rust
rust