Exploring the Potential of Reconstructed Multispectral Images for Urban Tree Segmentation in Street View Images

Keyphrases

• Urban Trees 100%
• Street View Images 100%
• Segmentation Network 100%
• Multispectral Image 100%
• Tree Segmentation 100%
• Near-infrared Image 50%
• Hyperspectral 50%
• Conditional Generative Adversarial Network (cGAN) 50%
• Near-infrared 50%
• Jekyll 50%
• No Significant Difference 25%
• T-test 25%
• Data Augmentation 25%
• Deep Learning 25%
• Spectral Bands 25%
• RGB Image 25%
• Deep Learning Model 25%
• Tree Identification 25%
• Specific Wavelength 25%
• Customs Data 25%
• Output Image 25%
• DeepLabv3+ 25%
• Cost-effective Solution 25%
• Hyperspectral Image Reconstruction 25%
• Captured Images 25%
• Infrared Information 25%
• SegFormer 25%
• Tree Dataset 25%
• Spectral Image 25%
• Red Edge 25%
• Specific Output 25%
• Structural Similarity Index 25%

Computer Science

• Multispectral Image 100%
• Network Segmentation 100%
• Hyperspectral Image 50%
• Generative Adversarial Networks 50%
• Deep Learning Model 25%
• Data Augmentation 25%
• Image Reconstruction 25%
• Captured Image 25%
• Effective Solution 25%
• Structural Similarity 25%
• Spectral Image 25%
• RGB Image 25%
• Deep Learning Method 25%

Earth and Planetary Sciences

• Multispectral Image 100%
• Hyperspectral Image 50%
• Image Reconstruction 25%
• Similarity Index 25%
• Spectral Band 25%