Multi-scale depth classification network for monocular depth estimation. (September 2022)

• Record Type: Journal Article
• Authors: Yang, Yi; Tian, Lihua; Li, Chen; Zhang, Botong
• Is Part Of: Computers & electrical engineering. Volume 102(2022)
• Publication Date: 2022-09
• Subjects: Depth estimation -- Classification -- Multi-scale -- Monocular vision; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2022.108206
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: A multiscale classification network for monocular depth estimation is proposed by transforming regression tasks into classification tasks. Depth map predicted by different resolution feature map contains different information. Depth estimation is carried out by depth classification layer at different scales. The final depth map is obtained by feature fusion of the multi-scaled depth map. Abstract: In addition to the RGB information of an image, depth information is the most critical. Monocular depth estimation is an effective method for predicting depth from RGB images. First, we propose a multiscale classification network that transforms the predicted depth values into a nonlinear combination problem between multiple depth interval values. Based on the correlation of the depth information, a depth classification network was used to determine the results of the weight values of each interval. Second, the depth maps predicted by feature maps with different resolutions contain different critical information. The lower-resolution depth maps play an overall role in predicting the overall contour; the higher-resolution depth map is more accurate in predicting object edge details. Finally, the desired depth map can be obtained by a 3 × 3 convolution of the predicted depth maps at multiple scales. We tested our proposed method on the NYU Depth V2 and KITTI datasets and achieved effective performance. Graphical abstract: Image, graphical abstract