Road marking extraction in UAV imagery using attentive capsule feature pyramid network

• Published in: International journal of applied earth observation and geoinformation Vol. 107; p. 102677
• Main Authors: Guan, Haiyan, Lei, Xiangda, Yu, Yongtao, Zhao, Haohao, Peng, Daifeng, Marcato Junior, José, Li, Jonathan
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.03.2022; Elsevier
• Subjects: Capsule; Dense atrous convolution; Feature pyramid network; neural networks; neurons; Road markings; spatial data; UAV images; Road markings; Capsule; Dense atrous convolution; Feature pyramid network; UAV images

•The proposed ACapsFPN extracted road markings in UAV images.•A capsule feature pyramid network provided a high-resolution, semantically-strong feature representation.•A novel multi-scale context feature (MCF) descriptor was designed to obtain multi-scale contextual information.•Ternary feature atte...