计算机科学
人工智能
深度学习
背景(考古学)
分割
融合
互补性(分子生物学)
卷积(计算机科学)
传感器融合
卷积神经网络
路面
RGB颜色模型
特征提取
人工神经网络
计算机视觉
模式识别(心理学)
机器学习
地理
工程类
生物
遗传学
哲学
土木工程
考古
语言学
作者
Antônio Edsom Carvalho Filho,Milton Hirokazu Shimabukuro,Aluir Porfírio Dal Póz
出处
期刊:IEEE Geoscience and Remote Sensing Letters
[Institute of Electrical and Electronics Engineers]
日期:2023-01-01
卷期号:20: 1-5
标识
DOI:10.1109/lgrs.2023.3291656
摘要
Road extraction is still a challenging topic for researchers. Currently, deep convolution neural networks are state-of-the-art in road network segmentation and are known for their remarkable ability to explore multilevel contexts. Despite this, the architectures still suffer from occlusion and obstructions that cause discontinuities and omissions in extracted road networks. Generally, these effects are minimized with strategies to obtain the context of the scene and not explore the complementarity of knowledge from a diversity of sources. We propose an early fusion network with RGB and surface model images that provide complementary geometric data to improve road surface extraction. Our results demonstrate that Unet_early reaches 71.01% IoU and 81.95% F1, and the fusion strategy increases the IoU and F1 scores at 2.3% and 1.5%, respectively. Besides, it overpassed the best model without fusion (DeepLabv3+). The Brazilian dataset and architecture implementation are available at https://github.com/tunofilho/ieee_road_multimodal.
科研通智能强力驱动
Strongly Powered by AbleSci AI