首页 > 论文 > 光学学报 > 37卷 > 10期(pp:1010003--1)

基于RGBD图像和卷积神经网络的快速道路检测

Fast Road Detection Based on RGBD Images and Convolutional Neural Network

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

针对移动平台有限的计算资源以及基于彩色图像的道路检测方法在极端光照情况下及路面类型变化时存在的不足, 提出了一种融合彩色图像和视差图像的基于9层卷积神经网络的快速道路检测算法。提出一种数据输入层预处理方法, 将视差图变换为视差梯度图以强化地面特征, 降低网络深度需求。所提两种网络结构为双通道后融合网络和单通道前融合网络, 分别用于卷积特征分析和快速道路检测。实验使用KITTI道路检测数据集并人为划分为普通和困难两组数据集, 对该算法进行实验对比和分析, 结果表明:与基于彩色图像的卷积神经网络方法相比, 该算法在普通数据集上最大F1指标(MaxF1)提升了1.61%, 在困难数据集上MaxF1提升了11.58%, 算法检测速度可达26 frame/s, 可有效克服光照、阴影、路面类型变化等影响。

Abstract

The road detection method based on color image exists problems under the extreme lighting conditions and changing road surface types, and the computing resource in moving platform is limited. So, based on the 9-layer convolutional neural network, a fast road detection algorithm is proposed to mix the color image and the disparity images. A new preprocessing method is applied in the data input layer, which can transform the disparity images to disparity gradient maps so as to enhance the representation of roads and reduce the demand for network depth. Two proposed networks are proposed including a double-path convolutional neural network which is used to analyze the characteristics of the convolutional neural network, and a single-path convolutional neural network which is applied to detect the road rapidly. The performance of the proposed algorithm is experimentally compared and analyzed on the KITTI road detection dataset which is divided into a common database and a difficult database artificially. The result demonstrates that, compared with the convolutional neural network method based on color images, the MaxF1 measures on the common database and difficult database improve by 1.61% and 11.58%, respectively, and the detection speed can be 26 frame/s. The proposed algorithm can overcome the impact of the lighting, shadow and the changing road surface effectively.

投稿润色
补充资料

中图分类号:TP391.4

DOI:10.3788/aos201737.1010003

所属栏目:图像处理

基金项目:国家自然科学基金(61671014,61601448)、上海市科技人才计划项目(14YF1407300)

收稿日期:2017-04-21

修改稿日期:2017-05-26

网络出版日期:--

作者单位    点击查看

曲 磊:中国科学院上海微系统与信息技术研究所仿生视觉系统实验室, 上海 200050中国科学院大学, 北京 100049
王康如:中国科学院上海微系统与信息技术研究所仿生视觉系统实验室, 上海 200050中国科学院大学, 北京 100049
陈利利:中国科学院上海微系统与信息技术研究所仿生视觉系统实验室, 上海 200050
李嘉茂:中国科学院上海微系统与信息技术研究所仿生视觉系统实验室, 上海 200050
张晓林:中国科学院上海微系统与信息技术研究所仿生视觉系统实验室, 上海 200050

联系人作者:曲磊(qulei@mail.sim.ac.cn)

备注:曲 磊(1989-), 男, 博士研究生, 主要从事计算机视觉方面的研究。

【1】Wang Wenfeng, Ding Weili, Li Yong, et al. An efficient road detection algorithm based on parallel edges[J]. Acta Optica Sinica, 2015, 35(7): 0715001.
王文锋, 丁伟利, 李 勇, 等. 一种高效的基于平行边缘的道路识别算法[J]. 光学学报, 2015, 35(7): 0715001.

【2】Duan Zhigang, Li Yong, Wang Ende, et al. Road and navigation line detection algorithm from shadow image based on the illumination invariant image[J]. Acta Optica Sinica, 2016, 36(12): 1215004.
段志刚, 李勇, 王恩德, 等. 基于光照不变图像的阴影图像道路及导航线提取算法[J]. 光学学报, 2016, 36(12): 1215004.

【3】Urtasun R, Lenz P, Geiger A. Are we ready for autonomous driving? The KITTI vision benchmark suite[C]. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2012: 3354-3361.

【4】Mendes C C T, Fr Mont V, Wolf D F. Exploiting fully convolutional neural networks for fast road detection[C]. IEEE International Conference on Robotics and Automation (ICRA), 2016: 3174-3179.

【5】Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[C]. International Conference on Neural Information Processing Systems (NIPS), 2012, 25(2): 1097-1105.

【6】Teichmann M, Weber M, Zoellner M, et al. MultiNet: real-time joint semantic reasoning for autonomous driving[J]. Computer Vision and Pattern Recognition, 2016.

【7】Oliveira G L, Burgard W, Brox T. Efficient deep models for monocular road segmentation[C]. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2016: 4885-4891.

【8】Gupta S, Girshick R, Arbel Ez P, et al. Learning rich features from RGB-D images for object detection and segmentation[C]. 13th European Conference on Computer Vision (ECCV), 2014: 345-360.

【9】Chen X Z, Ma H M, Wan J, et al. Multi-view 3D object detection network for autonomous driving[C]. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.

【10】Hu Z, Uchimura K. U-V-disparity: an efficient algorithm for stereovision based scene analysis[C]. IEEE Intelligent Vehicles Symposium (IVS), 2005: 48-54.

【11】Ojala T, Pietikainen M, Harwood D. Performance evaluation of texture measures with classification based on Kullback discrimination of distributions[C]. 12th International Conference on Pattern Recognition (ICPR), 1994: 582-585.

【12】Nair V, Hinton G E. Rectified linear units improve restricted boltzmann machines[C]. 27th International Conference on Machine Learning (ICML), 2010: 807-814.

【13】Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: a simple way to prevent neural networks from overfitting[J]. Journal of Machine Learning Research, 2014, 15(1): 1929-1958.

【14】Zbontar J, Lecun Y. Computing the stereo matching cost with a convolutional neural network[C]. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015: 1592-1599.

【15】Jia Y, Shelhamer E, Donahue J, et al. Caffe: convolutional architecture for fast feature embedding[J]. Proceedings of the 22nd ACM International Conference on Multimedia, 2014: 675-678.

引用该论文

Qu Lei,Wang Kangru,Chen Lili,Li Jiamao,Zhang Xiaolin. Fast Road Detection Based on RGBD Images and Convolutional Neural Network[J]. Acta Optica Sinica, 2017, 37(10): 1010003

曲 磊,王康如,陈利利,李嘉茂,张晓林. 基于RGBD图像和卷积神经网络的快速道路检测[J]. 光学学报, 2017, 37(10): 1010003

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF