Literature DB >> 34891097

Unsupervised learning of depth estimation from imperfect rectified stereo laparoscopic images.

Huoling Luo1, Congcong Wang2, Xingguang Duan3, Hao Liu4, Ping Wang5, Qingmao Hu1, Fucang Jia6.   

Abstract

BACKGROUND: Learning-based methods have achieved remarkable performances on depth estimation. However, the premise of most self-learning and unsupervised learning methods is built on rigorous, geometrically-aligned stereo rectification. The performances of these methods degrade when the rectification is not accurate. Therefore, we explore an approach for unsupervised depth estimation from stereo images that can handle imperfect camera parameters.
METHODS: We propose an unsupervised deep convolutional network that takes rectified stereo image pairs as input and outputs corresponding dense disparity maps. First, a new vertical correction module is designed for predicting a correction map to compensate for the imperfect geometry alignment. Second, the left and right images, which are reconstructed based on the input image pair and corresponding disparities as well as the vertical correction maps, are regarded as the outputs of the generative term of the generative adversarial network (GAN). Then, the discriminator term of the GAN is used to distinguish the reconstructed images from the original inputs to force the generator to output increasingly realistic images. In addition, a residual mask is introduced to exclude pixels that conflict with the appearance of the original image in the loss calculation.
RESULTS: The proposed model is validated on the publicly available Stereo Correspondence and Reconstruction of Endoscopic Data (SCARED) dataset and the average MAE is 3.054 mm.
CONCLUSION: Our model can effectively handle imperfect rectified stereo images for depth estimation.
Copyright © 2021 Elsevier Ltd. All rights reserved.

Entities:  

Keywords:  Depth estimation; Imperfect rectified stereo images; Laparoscopic image; Stereo matching; Unsupervised learning

Year:  2021        PMID: 34891097     DOI: 10.1016/j.compbiomed.2021.105109

Source DB:  PubMed          Journal:  Comput Biol Med        ISSN: 0010-4825            Impact factor:   4.589


  1 in total

1.  Joint estimation of depth and motion from a monocular endoscopy image sequence using a multi-loss rebalancing network.

Authors:  Shiyuan Liu; Jingfan Fan; Dengpan Song; Tianyu Fu; Yucong Lin; Deqiang Xiao; Hong Song; Yongtian Wang; Jian Yang
Journal:  Biomed Opt Express       Date:  2022-04-11       Impact factor: 3.562

  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.