Literature DB >> 33498363

A CNN-Based Length-Aware Cascade Road Damage Detection Approach.

Huiqing Xu1,2, Bin Chen2,3, Jian Qin4,5.   

Abstract

Accurate and robust detection of road damage is essential for public transportation safety. Currently, deep convolutional neural networks (CNNs)-based road damage detection algorithms to localize and classify damage with a bounding box have achieved remarkable progress. However, research in this field fails to take into account two key characteristics of road damage: weak semantic information and abnormal geometric properties, resulting in inappropriate feature representation and suboptimal detection results. To boost the performance, we propose a CNN-based cascaded damage detection network, called CrdNet. The proposed model has three parts: (1) We introduce a novel backbone network, named LrNet, that reuses low-level features and mixes suitable range dependency features to learn high-to-low level feature fusions for road damage weak semantic information representation. (2) We pan class="Gene">apply multi-scale and multiple aspect ratios anchor mechanism to generate high-quality positive samples regarding the damage with abnormal geometric properties for network training. (3) We designed an adaptive proposal assignment strategy and performed cascade predictions on corresponding branches that can establish different range dependencies. The experiments show that the proposed method achieves mean average precision (mAP) of 90.92% on a collected road damage dataset, demonstrating the good performance and robustness of the model.

Entities:  

Keywords:  abnormal geometric properties; deep convolutional neural network (CNN); length-aware; multi-scale attention; road damage detection

Year:  2021        PMID: 33498363      PMCID: PMC7864040          DOI: 10.3390/s21030689

Source DB:  PubMed          Journal:  Sensors (Basel)        ISSN: 1424-8220            Impact factor:   3.576


  5 in total

1.  DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs.

Authors:  Liang-Chieh Chen; George Papandreou; Iasonas Kokkinos; Kevin Murphy; Alan L Yuille
Journal:  IEEE Trans Pattern Anal Mach Intell       Date:  2017-04-27       Impact factor: 6.226

2.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks.

Authors:  Shaoqing Ren; Kaiming He; Ross Girshick; Jian Sun
Journal:  IEEE Trans Pattern Anal Mach Intell       Date:  2016-06-06       Impact factor: 6.226

3.  Deep High-Resolution Representation Learning for Visual Recognition.

Authors:  Jingdong Wang; Ke Sun; Tianheng Cheng; Borui Jiang; Chaorui Deng; Yang Zhao; Dong Liu; Yadong Mu; Mingkui Tan; Xinggang Wang; Wenyu Liu; Bin Xiao
Journal:  IEEE Trans Pattern Anal Mach Intell       Date:  2020-04-01       Impact factor: 6.226

4.  Res2Net: A New Multi-scale Backbone Architecture.

Authors:  Shanghua Gao; Ming-Ming Cheng; Kai Zhao; Xin-Yu Zhang; Ming-Hsuan Yang; Philip H S Torr
Journal:  IEEE Trans Pattern Anal Mach Intell       Date:  2019-08-30       Impact factor: 6.226

5.  Detection of Micro-Defects on Irregular Reflective Surfaces Based on Improved Faster R-CNN.

Authors:  Zhuangzhuang Zhou; Qinghua Lu; Zhifeng Wang; Haojie Huang
Journal:  Sensors (Basel)       Date:  2019-11-16       Impact factor: 3.576

  5 in total
  1 in total

1.  Image Sensing and Processing with Convolutional Neural Networks.

Authors:  Sonya Coleman; Dermot Kerr; Yunzhou Zhang
Journal:  Sensors (Basel)       Date:  2022-05-10       Impact factor: 3.847

  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.