.png){kind=icon}
.png){kind=icon}
Return
Automatic extraction of Martian dune crestlines under complex conditions: A comparative study of image processing, terrain analysis, and deep learning
.png){kind=icon}
PRE
AI
Y
Y
C
G
Z
DOI：10.1016/j.aeolia.2026.101047.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Digital Image Processing-based methods map crestline geometry but are noise-sensitive. • Digital Terrain Analysis-based methods ensure continuity but limit spatial accuracy. • Deep Learning-based methods show few false detections but suffer omissions and breaks. • Systematic comparison of three dune crestline extraction methods in Martian settings. • Error sources of the three methods are analyzed and specific improvements suggested.
Keywords:
Digital Image Processing
Digital Terrain Analysis
Deep Learning
Dune Crestline Extraction
Martian Surface Analysis
Journal
IF:
3.4
Papers:
782
Citations:
2.2K
Organization
Cited Papers