Wasserstein distance feature alignment learning for 2D image-based 3D model retrieval

Zhou, Yaqian; Liu, Yu; Zhou, Heyu; Li, Wenhui

doi:10.1016/j.jvcir.2021.103197

Cited by 8 publications

(2 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hojoon and Lee [8] abstracted the features of the model as nodes and then calculated their similarity to finalize the retrieval. Zhou et al [9] completed 3D model retrieval based on 2D image features. Li et al [10] proposed a multiview diagram matching method for 3D model retrieval, which aimed at breaking down and integrating the complex multiview diagram similarity degree measurement into several single view diagram similarity degree measurement.…”

Section: Introductionmentioning

confidence: 99%

3D Assembly Model Retrieval Method for a Multiassembly Interface

Wang

et al. 2022

Scientific Programming

View full text Add to dashboard Cite

3D model retrieval is increasingly becoming a hot topic in today’s research. The existing model retrieval methods are limited to the retrieval between similar models, while the matching retrieval for models with assembly relationships is neglected. They cannot realize the prediction of the virtual assembly and make the assemblers to spend extensive time detecting when there are assembly errors in the model. This work proposes a 3D assembly model retrieval method for multiassembly interface, which can retrieve the parts or subassemblies that can be assembled with the input model. In this work, a method for the reuse of the 3D assembly mode is also provided. The method here involves a series of steps, which includes an attribute adjacency graph to express the 3D assembly models, followed by the definition of a conjugated subgraph, and the serialization of the vertices of the graph for the Atlas of the assembly model. Finally, the frequent subgraph mining method is improved, and the 3D assembly model that satisfies the multiassembly interface is extracted. The article claims through the obtained results that this method not only allows the performance of the 3D assembly model retrieval of multiassembly interfaces but also the extraction of the 3D model required by designers, improving, at the same time, the product design efficiency.

show abstract

Section: Introductionmentioning

confidence: 99%

3D Assembly Model Retrieval Method for a Multiassembly Interface

Wang

et al. 2022

Scientific Programming

View full text Add to dashboard Cite

show abstract

“…Different methods have different emphases, but they all improve the retrieval effect of 3D models to varying degrees. Among them, the method based on color and shape distribution is widely used for its strong adaptability and wide application range ( Zhu et al., 2022 , Zhou et al., 2021 ). Kim ( Kim et al., 2017 ) proposed a shape distribution-based 3D CAD assembly comparison method, which can enable a comprehensive comparison of 3D CAD assembly models.…”

Section: Introductionmentioning

confidence: 99%

Feature extraction of 3D Chinese rose model based on color and shape features

et al. 2022

View full text Add to dashboard Cite

Flower classification is of great importance to the research fields of plants, food, and medicine. Due to more abundant information on three-dimensional (3D) flower models than two-dimensional 2D images, it makes the 3D models more suitable for flower classification tasks. In this study, a feature extraction and classification method were proposed based on the 3D models of Chinese roses. Firstly, the shape distribution method was used to extract the sharpness and contour features of 3D flower models, and the color features were obtained from the Red-Green-Blue (RGB) color space. Then, the RF-OOB method was employed to rank the extracted flower features. A shape descriptor based on the unique attributes of Chinese roses was constructed, χ2 distance was adopted to measure the similarity between different Chinese roses. Experimental results show that the proposed method was effective for the retrieval and classification tasks of Chinese roses, and the average classification accuracy was approximately 87%, which can meet the basic retrieval requirements of 3D flower models. The proposed method promotes the classification of Chinese roses from 2D space to 3D space, which broadens the research method of flower classification.

show abstract