Graph Convolution Algorithm Based on Visual Selectivity and Point Cloud Analysis Application

Abstract

The graph convolution algorithm currently suffers from the drawback of not fusing point cloud information and point cloud topology structure information based on visual selectivity features and using absolute quantities like distance as features, resulting in the algorithm losing geometric invariance. This information serves as the foundation for the "Graph Convolution Algorithm Based on Visual Selectivity and Application of Point Cloud Analysis". In order to propose a graph convolutional kernel and its design method based on visual selectivity, the algorithm analyzes the global characteristics of the point cloud "close in the vicinity and sparse in the distance," the local selectivity of the point cloud topology structure in the neighborhood, and the consistency between features and visual selectivity of primates. By combining point cloud information with point cloud topology structure information features, a graph convolution computation method was built, and the algorithm's geometric invariance was confirmed. The recognition and semantic segmentation performances of the approach in this study were verified using the ModelNet40 and ShapeNetPart data sets in comparison to the PointNet, PointNet++, DGCNN, KPConv, and 3D-GCN algorithms. The experimental design demonstrates that the algorithm presented in this research is accurate and practical, has geometric invariance, and performs better at semantic segmentation and recognition than conventional algorithms.

Adaptive Graph Convolution Algorithm Based on 3D Vision Selectivity and Its Application in Scene Segmentation

“…On the 𝑀𝑜𝑟𝑡𝑖𝑠𝑒_𝑎𝑛𝑑_𝑇𝑒𝑛𝑜𝑛_𝐷𝐵 dataset, using mIOU as the evaluation standard, and employing the network parameters set in Section 4.3, a comparative experiment was conducted between the MD-AGCNTDVS algorithm and traditional algorithms, and the comparative experiment results are shown in Figure 7. The results demonstrate that on the Mortise_and_Tenon_DB dataset, when compared with algorithms such as Kd-Net [34], MRTNet [34], PointNet [12], KCNet [22], RS-Net [23], SO-Net [26], PointNet++ [13], DGCNN [36], KPConv deform [25], 3D GCN [24], PCT [27] ,3D-Unet [28] and myalgorithm, our algorithm achieved segmentation results that are comparable to or better than traditional methods. This confirms that our algorithm is correct and feasible, and it has considerable algorithmic advantages.…”

“…Graph structure is an excellent way to learn and represent spatial structuring information. Thus, the graph convolution method has become a direction for 3D point cloud research [17][18][19][20], achieving certain research and application progress.…”

“…3D-GCN [24] uses a deformable 3D kernel to learn information from 3D point clouds and addresses the disorder and unstructured nature of point cloud data through graph-based max pooling methods, proposing a universal model concept. The GCN algorithm based on visual selectivity [25] proposes a method that combines point cloud information with point cloud topological structure features, establishing a graph convolution computation method. It effectively solves the problem of learning and integrating global and local features.…”

The Multi-View Deep Visual Adaptive Graph Convolution Network and Its Application in Point Cloud

“…In addition, the selected algorithms shown in Table 2 are highly representative in terms of point cloud processing, the reference algorithms include the rasterization method [5,[28][29][30], direct processing method [12,13], convolution processing method [33,34], graph convolution processing method [21], and graph convolution method based on visual computing [35]. The 3D point cloud processing technology was explained from several aspects and the experimental results indicate that these algorithms exhibited good classification performance on the dataset adopted in this study; the classification performance of the graph convolution algorithm was the best, the direct processing method performed well, followed by the early rasterization method.…”