Combination of modified U‐Net and domain adaptation for road detection. Issue 14 (1st December 2019)

• Record Type: Journal Article
• Authors: Dong, Ming; Zhao, Xiangmo; Fan, Xing; Shen, Chao; Liu, Zhanwen
• Is Part Of: IET image processing. Volume 13:Issue 14(2019)
• Page Start: 2735
• Page End: 2743
• Publication Date: 2019-12-01
• Subjects: learning (artificial intelligence) -- image classification -- object detection -- roads -- traffic engineering computing -- feature extraction -- road vehicles -- image segmentation -- computer vision; modified U‐Net -- scene understanding -- autonomous driving -- deep learning -- visual road detection problem -- complex illumination encounter -- detection model -- deep network architecture -- network U‐Net‐prior -- domain adaptation model -- DAM -- modified segmentation network -- training images -- test images -- adversarial learning -- monocular‐vision‐based methods; Image processing -- Periodicals; 621.36705
• DOI: 10.1049/iet-ipr.2018.6696
• Languages: English
• ISSNs: 1751-9659

Abstract

Abstract : Road detection is one of the crucial tasks for scene understanding in autonomous driving. Recently, methods based on deep learning had rapidly grown and addressed this task excellently, because they can extract more abundant features. In this study, the authors consider the visual road detection problem as a classification for each pixel of the given image, which is road or non‐road. There is complex illumination encounter in traffic applications, so that the detection model has poor adaptability. They address this problem by proposing a deep network architecture, which combines the network U‐Net‐prior and domain adaptation model (DAM). U‐Net‐prior is a modified segmentation network which integrates location prior and shape prior into U‐Net. DAM is a model for reducing the gap between training images and test images, which is optimised in adversarial learning to make the features extracted from different datasets close to each other. They validate the effectiveness of each component of the algorithm, and compare the overall architecture with other state‐of‐the‐art methods, and the results show that the architecture achieves top accuracies with the shortest run time in monocular‐vision‐based methods, simultaneously, compared with the methods based on other sensors, the architecture also achieves a competitive result.