Open AccessNew Method of Airport Pavement Health Inspection Based on MobileNet-SSD and Mask R-CNN
Author(s) -
Wentong Guo,
Hongyuan Fang,
Niannian Wang
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1885/2/022048
Subject(s) - pixel , computer science , object (grammar) , identification (biology) , segmentation , computer vision , artificial intelligence , object detection , measure (data warehouse) , image segmentation , pattern recognition (psychology) , data mining , botany , biology
This paper presents a rapid detection and pixel size measurement method for airport pavement apparent disease and Foreign Object Debris. Firstly, MobileNet-SSD algorithm is used for object identification. Then Mask R-CNN algorithm is used to segment the target image and measure the pixel size of the target object. Finally, the detailed information of the target object is obtained. The experimental verification shows that the recognition speed of this method reaches 65 frames per second, and the pixel segmentation accuracy reaches 96 %, which can meet the requirements of airport pavement health inspection.