<i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning
This paper aims to explore and evaluate aerial imagery and deep learning technology in pavement condition evaluation. A convolutional neural network (CNN) model, named <i>PCIer</i>, was designed to process aerial images and produce pavement condition index (PCI) estimations, which are cl...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Geographies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-7086/3/1/8 |
| _version_ | 1797611531371806720 |
|---|---|
| author | Sisi Han In-Hun Chung Yuhan Jiang Benjamin Uwakweh |
| author_facet | Sisi Han In-Hun Chung Yuhan Jiang Benjamin Uwakweh |
| author_sort | Sisi Han |
| collection | DOAJ |
| description | This paper aims to explore and evaluate aerial imagery and deep learning technology in pavement condition evaluation. A convolutional neural network (CNN) model, named <i>PCIer</i>, was designed to process aerial images and produce pavement condition index (PCI) estimations, which are classified into four scales of Good (PCI ≥ 70), Fair (50 ≤ PCI < 70), Poor (25 ≤ PCI < 50), and Very Poor (PCI < 25). In the experiment, the PCI datasets were retrieved from the published pavement condition report by the City of Sacramento, CA. Following the retrieved datasets, the authors also collected the corresponding aerial image datasets containing 100 images for each PCI grade from Google Earth. An 80% proportion of datasets were used for <i>PCIer</i> model training, and the remaining were used for testing. Comparisons showed using a 128-channel heatmap layer in the proposed <i>PCIer</i> model and saving the <i>PCIer</i> model with the best validation accuracy would yield the best performance, with a testing accuracy of 0.97, and a weighted average precision, recall, and F1-score of 0.98, 0.97, and 0.97, respectively. Moreover, future research recommendations are provided in the discussion for improving the effectiveness of pavement evaluation via aerial imagery and deep learning. |
| first_indexed | 2024-03-11T06:29:03Z |
| format | Article |
| id | doaj.art-60b9989d12b14c068895de405bce021d |
| institution | Directory Open Access Journal |
| issn | 2673-7086 |
| language | English |
| last_indexed | 2024-03-11T06:29:03Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Geographies |
| spelling | doaj.art-60b9989d12b14c068895de405bce021d2023-11-17T11:19:21ZengMDPI AGGeographies2673-70862023-02-013113214210.3390/geographies3010008<i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep LearningSisi Han0In-Hun Chung1Yuhan Jiang2Benjamin Uwakweh3Department of Civil, Construction and Environmental Engineering, Marquette University, Milwaukee, WI 53233, USADepartment of Construction and Operations Management, South Dakota State University, Brookings, SD 57007, USADepartment of Built Environment, North Carolina A&T State University, Greensboro, NC 27411, USADepartment of Built Environment, North Carolina A&T State University, Greensboro, NC 27411, USAThis paper aims to explore and evaluate aerial imagery and deep learning technology in pavement condition evaluation. A convolutional neural network (CNN) model, named <i>PCIer</i>, was designed to process aerial images and produce pavement condition index (PCI) estimations, which are classified into four scales of Good (PCI ≥ 70), Fair (50 ≤ PCI < 70), Poor (25 ≤ PCI < 50), and Very Poor (PCI < 25). In the experiment, the PCI datasets were retrieved from the published pavement condition report by the City of Sacramento, CA. Following the retrieved datasets, the authors also collected the corresponding aerial image datasets containing 100 images for each PCI grade from Google Earth. An 80% proportion of datasets were used for <i>PCIer</i> model training, and the remaining were used for testing. Comparisons showed using a 128-channel heatmap layer in the proposed <i>PCIer</i> model and saving the <i>PCIer</i> model with the best validation accuracy would yield the best performance, with a testing accuracy of 0.97, and a weighted average precision, recall, and F1-score of 0.98, 0.97, and 0.97, respectively. Moreover, future research recommendations are provided in the discussion for improving the effectiveness of pavement evaluation via aerial imagery and deep learning.https://www.mdpi.com/2673-7086/3/1/8aerial imageryconvolutional neural network (CNN)pavement condition index (PCI) |
| spellingShingle | Sisi Han In-Hun Chung Yuhan Jiang Benjamin Uwakweh <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning Geographies aerial imagery convolutional neural network (CNN) pavement condition index (PCI) |
| title | <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning |
| title_full | <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning |
| title_fullStr | <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning |
| title_full_unstemmed | <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning |
| title_short | <i>PCIer</i>: Pavement Condition Evaluation Using Aerial Imagery and Deep Learning |
| title_sort | i pcier i pavement condition evaluation using aerial imagery and deep learning |
| topic | aerial imagery convolutional neural network (CNN) pavement condition index (PCI) |
| url | https://www.mdpi.com/2673-7086/3/1/8 |
| work_keys_str_mv | AT sisihan ipcieripavementconditionevaluationusingaerialimageryanddeeplearning AT inhunchung ipcieripavementconditionevaluationusingaerialimageryanddeeplearning AT yuhanjiang ipcieripavementconditionevaluationusingaerialimageryanddeeplearning AT benjaminuwakweh ipcieripavementconditionevaluationusingaerialimageryanddeeplearning |