This paper is inspired by the U-Net semantic segmentation network and holistically nested edge detection network.
This creates difficulties in the image detection task. In practical applications, surface cracks are divided into inner and edge regions for pavements with rough surfaces and complex environments. In recent years, deep learning-based detection methods have been applied to pavement crack detection.
Asphalt depth gauge Patch#
A set of sand patch field tests are implemented to verify the reliability of the developed method. According to the profile method, an integrated calculation procedure of texture depth is established. Through the model and the spatial equations obtained, the values of the mean profile depth of asphalt pavement can be figured out. In addition, the techniques of multi-line laser pairing and epipolar restriction are introduced for the image matching between multi-line laser and binocular vision. Image processing technology helps locate the exact coordinates of each point on the model. A three-dimensional mathematical model is established based on the triangulation measuring principle in order to implement the three-dimensional reconstruction of asphalt surface profile. The developed method is incorporated with the profile method and the digital image technology. A new measuring method of mean texture depth of asphalt pavement is proposed based on the technology of multi-line laser and binocular vision.
0 kommentar(er)
