Body size measurement based on deep learning for image segmentation by binocular stereovision system

Song, Xiaowei; Song, Xiaoguo; Yang, Lei; Li, Menglong; Hou, Chunping; Xiong, Zixiang

doi:10.1007/s11042-021-11470-2

Cited by 2 publications

(10 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The process of the stereovision-based anthropometry can be divided into three steps: semantic segmentation of the girth region; stereo matching and coordinate calculation; and girth fitting [ 17 , 18 ]. The flowchart is shown in Figure 1 .…”

Section: The Proposed Methodsmentioning

confidence: 99%

“…In the anthropometric system in reference [ 17 , 18 ], color markers are used for stereo matching, and the matching point pair closest to the center of the marker is reserved for spatial coordinate calculation. In the anthropometric system in this paper, corners are used for stereo matching.…”

Section: The Proposed Methodsmentioning

confidence: 99%

“…The anthropometric methods in [ 17 , 18 ] make a trade-off between the accuracy and complexity of the aforementioned two types of anthropometric methods with optical cameras. In reference [ 17 ], three pairs of synchronously triggered stereo cameras were adopted to collect three pairs of stereo images from the front, side and back of a subject.…”

Section: Introductionmentioning

confidence: 99%

“…In reference [ 17 ], three pairs of synchronously triggered stereo cameras were adopted to collect three pairs of stereo images from the front, side and back of a subject. In reference [ 18 ], one pair of stereo cameras and a turntable were used to acquire four pairs of stereo images of a subject from four different views with partially overlapping areas. Both methods made use of the 3D spatial information obtained through stereo matching and coordinate calculation of markers to improve the measurement accuracies, which are greater than those of the 2D-image-processing-based methods in [ 11 , 12 , 13 ].…”

Section: Introductionmentioning

confidence: 99%

“…The matching precision is refined to the sub-pixel level by the checkerboard corner design and the corresponding corner detection algorithm, and the matching complexity is reduced by the regional constraint of the checkerboard corner. The proposed algorithm and design can significantly improve the accuracy of the anthropometric system in [ 17 , 18 ].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

An Improved Human-Body-Segmentation Algorithm with Attention-Based Feature Fusion and a Refined Stereo-Matching Scheme Working at the Sub-Pixel Level for the Anthropometric System

Yang

Guo

Song

et al. 2022

Entropy

Self Cite

View full text Add to dashboard Cite

This paper proposes an improved human-body-segmentation algorithm with attention-based feature fusion and a refined corner-based feature-point design with sub-pixel stereo matching for the anthropometric system. In the human-body-segmentation algorithm, four CBAMs are embedded in the four middle convolution layers of the backbone network (ResNet101) of PSPNet to achieve better feature fusion in space and channels, so as to improve accuracy. The common convolution in the residual blocks of ResNet101 is substituted by group convolution to reduce model parameters and computational cost, thereby optimizing efficiency. For the stereo-matching scheme, a corner-based feature point is designed to obtain the feature-point coordinates at sub-pixel level, so that precision is refined. A regional constraint is applied according to the characteristic of the checkerboard corner points, thereby reducing complexity. Experimental results demonstrated that the anthropometric system with the proposed CBAM-based human-body-segmentation algorithm and corner-based stereo-matching scheme can significantly outperform the state-of-the-art system in accuracy. It can also meet the national standards GB/T 2664-2017, GA 258-2009 and GB/T 2665-2017; and the textile industry standards FZ/T 73029-2019, FZ/T 73017-2014, FZ/T 73059-2017 and FZ/T 73022-2019.

show abstract

Section: The Proposed Methodsmentioning

confidence: 99%

Section: The Proposed Methodsmentioning

confidence: 99%