Method for Estimating Monocular Image Depth Based on Dense Convolutional Network

doi:10.19678/j.issn.1000-3428.0059516

Abstract

Abstract: To address the low accuracy and complex network structure of the existing methods for estimating monocular image depth,a dense convolutional network for estimating monocular image depth is proposed,which adopts an end-to-end encoder and decoder.Dense convolutional Network(DenseNet) is introduced into the encoder,and the output of each previous layer is taken as the input of this layer,which enhances feature reuse and forward propagation while reducing the number of parameters and network computation,thus avoiding the occurrence of gradient disappearance to a certain extent.The decoder adopts the upper projection module with cavity convolution and the bilinear upper sampling module to better express the image features extracted by the encoder,and finally obtain the estimated depth map corresponding to the input image.The proposed network is trained,verified and tested on NYU Depth V2,an indoor scene depth data set.The results show that the proposed dense convolutional network structure achieves an accuracy of 0.851 and a Root Mean Square Error(RMSE) of 0.482 in the case of δ<1.25.

Key words: dense convolutional network, monocular image, encoder, decoder, depth estimation

摘要： 为解决目前单目图像深度估计方法存在的精度低、网络结构复杂等问题，提出一种密集卷积网络结构，该网络采用端到端的编码器和解码器结构。编码器引入密集卷积网络DenseNet，将前面每一层的输出作为本层的输入，在加强特征重用和前向传播的同时减少参数量和网络计算量，从而避免梯度消失问题发生。解码器结构采用带有空洞卷积的上投影模块和双线性插值模块，以更好地表达由编码器所提取的图像特征，最终得到与输入图像相对应的估计深度图。在NYU Depth V2室内场景深度数据集上进行训练、验证和测试，结果表明，该密集卷积网络结构在δ<1.25时准确率达到0.851，均方根误差低至0.482。

关键词: 密集卷积网络, 单目图像, 编码器, 解码器, 深度估计

CLC Number:

TP391

WANG Yaqun, DAI Hualin, WANG Li, LI Guoyan. Method for Estimating Monocular Image Depth Based on Dense Convolutional Network[J]. Computer Engineering, 2021, 47(11): 262-267,291.

王亚群, 戴华林, 王丽, 李国燕. 基于密集卷积网络的单目图像深度估计方法[J]. 计算机工程, 2021, 47(11): 262-267,291.

/ / Recommend / Download Citations

URL: http://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0059516

http://www.ecice06.com/EN/Y2021/V47/I11/262

Figures/Tables 12

References

[1] 曾仕峰,吴锦均,叶智文,等.基于ROS的无人驾驶智能车[J].物联网技术,2020,10(6):62-63,66. ZENG S F,WU J J,YE Z W,et al.ROS based driverless intelligent car[J].Internet of Things Technologies,2020,10(6):62-63,66.(in Chinese)
[2] 赵妍,解迎刚,陈莉莉.基于SLAM算法的机器人智能激光定位技术的研究[J].激光杂志,2019,40(7):169-173. ZHAO Y,XIE Y G,CHEN L L.Research on robot intelligent laser localization based on SLAM algorithm[J].Laser Journal,2019,40(7):169-173.(in Chinese)
[3] LIU X T,SINHA A,ISHII M,et al.Dense depth estimation in monocular endoscopy with self-supervised learning methods[J].IEEE Transactions on Medical Imaging,2019,39(5):1438-1447.
[4] ELKERDAWY S,ZHANG H,RAY N.Lightweight monocular depth estimation model by joint end-to-end filter pruning[C]//Proceedings of International Conference on Image Processing.Washington D.C.,USA:IEEE Press,2019:4290-4294.
[5] WOFK D,FANGCHANG M A,YANG T J,et al.Fast depth:fast monocular depth estimation on embedded systems[C]//Proceedings of International Conference on Robotics and Automation.Washington D.C.,USA:IEEE Press,2019:6101-6108.
[6] POGGI M,ALEOTTI F,TOSI F,et al.Towards real-time unsupervised monocular depth estimation on CPU[C]//Proceedings of RSJ International Conference on Intelligent Robots and Systems.Washington D.C.,USA:IEEE Press,2018:5848-5854.
[7] EIGEN D,FERGUS R.Predicting depth,surface normals and semantic labels with a common multi-scale convolutional architecture[C]//Proceedings of International Conference on Computer Vision.Washington D.C.,USA:IEEE Press,2015:2650-2658.
[8] LIU F Y,SHEN C H,LIN G S,et al.Learning depth from single monocular images using deep convolutional neural fields[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(10):2024-2039.
[9] LAINA I,RUPPRECHT C,BELAGIANNIS V,et al.Deeper depth prediction with fully convolutional residual networks[C]//Proceedings of 2016 International Conference on 3D Vision.Washington D.C.,USA:IEEE Press,2016:239-248.
[10] TEED Z,DENG J.Deepv2d:video to depth with differentiable structure from motion[EB/OL].[2020-08-05].https://arxiv.org/pdf/1812.04605.pdf.
[11] GODARD C,MAC AODHA O,FIRMAN M,et al.Digging into self-supervised monocular depth estimation[C]//Proceedings of International Conference on Computer Vision.Washington D.C.,USA:IEEE Press,2019:3828-3838.
[12] YIN W,LIU Y F,SHEN C H,et al.Enforcing geometric constraints of virtual normal for depth prediction[C]//Proceedings of International Conference on Computer Vision.Washington D.C.,USA:IEEE Press,2019:5684-5693.
[13] LEE J H,HAN M K,KO D W,et al.From big to small:multi-scale local planar guidance for monocular depth estimation[EB/OL].[2020-08-05].https://arxiv.org/pdf/1907.10326v1.pdf.
[14] TOSI F,ALEOTTI F,POGGI M,et al.Learning monocular depth estimation infusing traditional stereo knowledge[C]//Proceedings of Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2019:9799-9809.
[15] GUIZILINI V,LI J,AMBRUS R,et al.Robust semi-supervised monocular depth estimation with reprojected distances[C]//Proceedings of Conference on Robot Learning.Washington D.C.,USA:IEEE Press,2020:503-512.
[16] HUANG G,LIU Z,VAN DER MAATEN L,et al.Densely connected convolutional networks[C]//Proceedings of Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2017:4700-4708.
[17] YU F,KOLTUN V.Multi-scale context aggregation by dilated convolutions[EB/OL].[2020-08-05].https://arxiv.org/pdf/1511.07122.pdf.
[18] HU J J,OZAY M,ZHANG Y,et al.Revisiting single image depth estimation:toward higher resolution maps with accurate object boundaries[C]//Proceedings of Winter Conference on Applications of Computer Vision.Washington D.C.,USA:IEEE Press,2019:1043-1051.
[19] IOFFE S,SZEGEDY C.Batch normalization:accelerating deep network training by reducing internal covariate shift[EB/OL].[2020-08-05].http://proceedings.mlr.press/v37/ioffe15.pdf.
[20] HINTON G E,SRIVASTAVA N,KRIZHEVSKY A,et al.Improving neural networks by preventing co-adaptation of feature detectors[J].Computer Ence,2012,3(4):212-223.
[21] 王泉德,张松涛.基于多尺度特征融合的单目图像深度估计[J].华中科技大学学报(自然科学版),2020,48(5):7-12. WANG Q D,ZHANG S T.Monocular image depth estimation based on multi-scale feature fusion[J].Journal of Huazhong University of Science and Technology (Natural Science Edition),2020,48(5):7-12.(in Chinese)
[22] CHEN L C,PAPANDREOU G,SCHROFF F,et al.Rethinking atrous convolution for semantic image segmentation[EB/OL].[2020-08-05].https://arxiv.org/pdf/1706.05587.pdf.
[23] RICCI E,OUYANG W L,WANG X G,et al.Monocular depth estimation using multi-scale continuous CRFs as sequential deep networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,41(6):1426-1440.

Please choose a citation manager

Content to export