The 3D virtual mannequin has been widely used in apparel industry, and its importance is also increasing. This work develops a new 3D virtual mannequin reconstruction system based on optimization. All the mannequins reconstructed by the proposed approach share the identical topology, that is, there is a point-to-point correspondence among the mannequins, which will significantly facilitate much subsequent processing in fashion design, made-to-measure, and virtual try-on. The inputs to the proposed system contain a template human body, a raw scan (represented in mesh), and a very sparse corresponding landmarks set. The proposed approach substantially utilizes the optimization technology to drive the template to deform into a real scan. There is no special requirement on the raw meshes. The raw meshes may have a different number of vertices and triangles or may even be incomplete. The proposed method only needs 21 landmarks as hard-constraints to reconstruct a mannequin with tens of thousands of vertices. These landmarks can be extracted automatically for standard mannequin reconstruction. Besides the standard mannequin, the proposed system can also be used to reconstruct display mannequins, that is, mannequins with various poses. The experiments visualize the optimization procedure and verify that the optimization is efficient and effective. Quantitative analysis also proves that the reconstruction error satisfies the requirements of fashion design and tailoring.

In this paper, we present a novel non-parametric method for precisely reconstructing a three dimensional (3D) virtual mannequin from anthropometric measurements and mask image(s) based on Graph Convolution Network (GCN). The proposed method avoids heavy dependence on a particular parametric body model such as SMPL or SCPAE and can predict mesh vertices directly, which is significantly more comfortable using a GCN than a typical Convolutional Neural Network (CNN). To further improve the accuracy of the reconstruction and make the reconstruction more controllable, we incorporate the anthropometric measurements into the developed GCN. Our non-parametric reconstruction results distinctly outperform the previous graph convolution method, both visually and in terms of anthropometric accuracy. We also demonstrate that the proposed network possesses the capability to reconstruct a plausible 3D mannequin from a single-view mask. The proposed method can be effortless extended to a parametric method by appending a Multilayer Perception (MLP) to regress the parametric space of the Principal Component Analysis (PCA) model to achieve 3D reconstruction as well. Extensive experimental results demonstrate that our anthropometric GCN itself is very useful in improving the reconstruction accuracy, and the proposed method is effective and robust for 3D mannequin reconstruction.

Automatic 3D human body measurement is a crucial issue for tailoring and made-to-measure. This paper presents a novel framework for 3D human landmarks extraction and measurement. The proposed approach first segments the 3D human body into 13 parts by utilizing an improved Mean Curvature Skeleton (MCS) algorithm, in which we modify the Laplacian operator used in the original MCS with mesh saliency to enable the segmentation boundaries to be closer to the human joints. Based on the human segmentation, K-Nearest Neighbors, linear modeling, and geometric methods are employed to extract at least 21 landmarks. Many essential landmarks, such as acromion, elbow, crotch, etc., are extracted in new ways. To our best knowledge, this is the first paper to propose a generalized solution to approximate the elbow point for arbitrary arm poses in the automatic 3D human measurement systems. Subsequently, various anthropometric measurements can be calculated according to the landmarks extracted automatically. The proposed method is validated on the public datasets and our real scans, and the experimental results have verified that the proposed approach is efficient and effective in processing various 3D human bodies.

The detection of unsound wheat kernels in traditional wheat purchasing is affected by human factors, resulting in wrong wheat grading. At present, computer-based recognition of wheat kernels has generally low accuracy, and few types of wheat kernels can be recognized. To quickly, accurately, and objectively recognize wheat kernels, this study proposed an improved strategy of wheat kernel recognition method based on deep learning. First, a large number of collected wheat images were labeled, and the wheat kernels were divided into five categories: perfect kernels, broken kernels, impurities, sprouted kernels, and moldy kernels. Second, the improved strategies of VggNet-16, ResNet-34, EfficientNet-b2, DenseNet121, and Vit models were proposed. Based on the two-stage target detection method, the improved network model was used to detect wheat kernels. Moreover, the accuracy of the model was verified by performing comparative tests. Results show that the improved network structure is obviously improved, and the highest accuracy rate of wheat kernel identification is 96%. The precision, recall rate, and F1-score of VggNet-16-W, ResNet-34-W, EfficientNet-b2-W, and DenseNet121-W models are above 97%. This study provides a good reference for rapid and accurate detection of wheat quality. Keywords: deep learning; image recognition; improved strategies; network model; wheat kernels

In order to generate the three-dimensional (3D) mesh of draped fabric based on a two-dimensional (2D) fabric projection, the 3D point cloud of draped fabric was scanned via a self-built 3D scanning device followed by triangulation. The 3D boundary of fabric drape model and the two-dimensional contour of the projection of the fabric drape model were extracted respectively. Two neural networks were constructed to bridge the 2D fabric drape projection and the 3D boundary of draped fabric. With the trained neural networks, the 3D boundary of draped fabric could be inferred by merely referring a 2D projection of draped fabric. Six reference triangular meshes with different triangle density were generated. Driven by two mesh deformation methods, Laplacian deformation and Poisson deformation, the reference mesh was deformed into 3D mesh similar to scanned fabric drape models (meshes). The result shows that the generated models have a high agreement with the scanned meshes. In addition, the Laplacian deformation outperforms Poisson deformation in generating 3D fabric drape models.

In this paper, the problem of inferring fabric mechanical properties based on three-dimensional (3D) fabric drape model and fabric weight was studied. Firstly, the three-dimensional point cloud of draped fabric was scanned via a self-built three-dimensional scanning device followed with triangulation. Then the local linear embedding (LLE) algorithm was used to resample the three-dimensional draped models (triangular meshes). Besides, all meshes resampled with the same uniform discrete points have the same triangular topology (vertex number and triangle index). Thirdly, the distribution function of curvature (DFC) of resampled 3 D mesh was extracted. At last, four neural networks with the combination of fabric weight and DFC as input were constructed to predict the bending rigidity and shearing stiffness separately. The result shows that the correlation coefficient between true warp bending rigidity and predicted result could reach 0.925. The correlation coefficient between true warp shearing stiffness and the predicted result is 0.7953. The correlation coefficient between true weft bending rigidity and the predicted result is 0.857. The correlation coefficient between true weft shearing stiffness and the predicted result is 0.605.