A Neural Network Approach to Semantic Segmentation of Vehicles in Very High Resolution Images

View Full Paper

Ask AI

Bookmark

The proposed neural network method achieves over 90% segmentation accuracy, indicating significant effectiveness.
Utilizing the DeepLab architecture with ResNet-101 enhances multiscale feature analysis, which preserves object texture.
The study applies a loss function based on the Dice coefficient to effectively address class imbalance issues.
Results outperform existing methods, achieving a mean Intersection over Union (mIoU) of 82.3% and pixel accuracy of 95.1%.

Discussion

No takes yet. Share an insight, caveat, or question.

This approach shows improved segmentation accuracy in analyzing urban environments, suggesting viable applications in traffic management.

Kashtan et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark