Non-local affinity adaptive acceleration propagation network for generating dense depth maps from LiDAR

被引:0
|
作者
Zhang, Hanxuan [1 ]
Huo, Ju [1 ]
机构
[1] Harbin Inst Technol, Sch Elect Engn & Automat, Harbin 150000, Peoples R China
基金
中国国家自然科学基金;
关键词
COMPLETION; MODEL; NET;
D O I
10.1364/OE.492187
中图分类号
O43 [光学];
学科分类号
070207 ; 0803 ;
摘要
Depth completion aims to generate dense depth maps from the sparse depth images generated by LiDAR. In this paper, we propose a non-local affinity adaptive accelerated (NL-3A) propagation network for depth completion to solve the mixing depth problem of different objects on the depth boundary. In the network, we design the NL-3A prediction layer to predict the initial dense depth maps and their reliability, non-local neighbors and affinities of each pixel, and learnable normalization factors. Compared with the traditional fixed-neighbor affinity refinement scheme, the non-local neighbors predicted by the network can overcome the propagation error problem of mixed depth objects. Subsequently, we combine the learnable normalized propagation of non-local neighbor affinity with pixel depth reliability in the NL-3A propagation layer, so that it can adaptively adjust the propagation weight of each neighbor during the propagation process, which enhances the robustness of the network. Finally, we design an accelerated propagation model. This model enables parallel propagation of all neighbor affinities and improves the efficiency of refining dense depth maps. Experiments on KITTI depth completion and NYU Depth V2 datasets show that our network is superior to most algorithms in terms of accuracy and efficiency of depth completion. In particular, we predict and reconstruct more smoothly and consistently at the pixel edges of different objects.
引用
收藏
页码:22012 / 22029
页数:18
相关论文
共 33 条
  • [1] ADNet: Non-Local Affinity Distillation Network for Lightweight Depth Completion With Guidance From Missing LiDAR Points
    Kim, Janghyun
    Noh, Jeonghyun
    Jeong, Mingyu
    Lee, Wonju
    Park, Yeonchool
    Park, Jinsun
    IEEE ROBOTICS AND AUTOMATION LETTERS, 2024, 9 (09): : 7533 - 7540
  • [2] SGTBN: Generating Dense Depth Maps From Single-Line LiDAR
    Lu, Hengjie
    Xu, Shugong
    Cao, Shan
    IEEE SENSORS JOURNAL, 2021, 21 (17) : 19091 - 19100
  • [3] DAN-Conv: Depth aware non-local convolution for LiDAR depth completion
    Yan, Lin
    Liu, Kai
    Gao, Long
    ELECTRONICS LETTERS, 2021, 57 (20) : 754 - 757
  • [4] Non-local Haze Propagation with an Iso-Depth Prior
    Kim, Incheol
    Kim, Min H.
    COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS - THEORY AND APPLICATIONS, VISIGRAPP 2017, 2019, 983 : 213 - 238
  • [5] Adaptive colour-guided non-local means algorithm for compound noise reduction of depth maps
    Ibrahim, Mostafa M.
    Liu, Qiong
    Yang, You
    IET IMAGE PROCESSING, 2020, 14 (12) : 2768 - 2779
  • [6] LiStereo: Generate Dense Depth Maps from LIDAR and Stereo Imagery
    Zhang, Junming
    Ramanagopal, Manikandasriram Srinivasan
    Vasudevan, Ram
    Johnson-Roberson, Matthew
    2020 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), 2020, : 7829 - 7836
  • [7] Adaptive Non-local Affinity Graph for Unsupervised Image Segmentation
    Lv, Xin
    Su, Zhenming
    Zhang, Taiyi
    Cheng, Wenxiang
    Qi, Xiaoqiong
    2023 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, ICME, 2023, : 2357 - 2362
  • [8] GENERATING DENSE DEPTH MAPS USING A PATCH CLOUD AND LOCAL PLANAR SURFACE MODELS
    Herrera, Daniel C.
    Kannala, Juho
    Heikkil, Janne
    2011 3DTV CONFERENCE: THE TRUE VISION - CAPTURE, TRANSMISSION AND DISPLAY OF 3D VIDEO (3DTV-CON), 2011,
  • [9] NLDN: Non-local dehazing network for dense haze removal
    Zhang, Shengdong
    He, Fazhi
    Ren, Wenqi
    NEUROCOMPUTING, 2020, 410 (410) : 363 - 373
  • [10] MCAR: Non-local adaptive Network-on-Chip routing with message propagation of congestion information
    Xie, Ruilian
    Cai, Jueping
    Xin, Xin
    Yang, Bo
    MICROPROCESSORS AND MICROSYSTEMS, 2017, 49 : 117 - 126