A context-adaptive fusion of shadow and highlight classifiers reports 96.75% accuracy on a real sonar image set, alongside a new synthetic naval-mine dataset with physics-informed noise.
Random Forest and Linear Regression have been used to distinguish rocks from mines [4]
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
A Novel Context-Adaptive Fusion of Shadow and Highlight Regions for Efficient Sonar Image Classification
A context-adaptive fusion of shadow and highlight classifiers reports 96.75% accuracy on a real sonar image set, alongside a new synthetic naval-mine dataset with physics-informed noise.