Due to the significant bandwidth and memory requirements for transmitting and storing large-scale point clouds, considerable progress has been made in recent years in the field of large-scale point cloud geometry compression. However, challenges remain, including suboptimal compression performance and complex encoding–decoding processes. To address these issues, we propose an efficient large-scale scene point cloud geometry compression algorithm. By analyzing the sparsity of large-scale point clouds and the impact of scale on feature extraction, we design a cross-attention module in the encoder to enhance the extracted features by incorporating positional information. During decoding, we introduce an efficient generation module that improves decoding quality without increasing decoding time. Experiments on three public datasets demonstrate that, compared to the state-of-the-art G-PCC v23, our method achieves an average bitrate reduction of −46.64%, the fastest decoding time, and a minimal network model size of 2.8 M.

Three-dimensional (3D) registration plays a pivotal step in augmented reality (AR) systems. Traditional 3D registration methods have the disadvantages of poor accuracy and robustness. This paper proposes a novel registration method, we believe, for AR systems based on the AKAZE and Tanimoto similarity measurement method. In this paper, the image feature points are extracted and matched by combining the AKAZE algorithm and the Tanimoto similarity measurement method. Then, the camera pose is estimated by calculating the constraint relationship of the feature points. Finally, the 3D registration and real-time tracking of the virtual objects are realized by the Lucas–Kanade (LK) optical flow tracking algorithm. We use Tanimoto to determine the similarity of feature points to improve the matching accuracy of the AKAZE algorithm, and this method not only retains the advantages of strong scale adaptation but also has the advantage of high-precision matching. Experiments show that the method proposed in this paper has the benefits of high registration accuracy, low time consumption, and strong robustness. Under the premise of ensuring accuracy, when the marker is rotated or blocked, it can be accurately registered. In addition, when the external environment changes, for example, the light intensity or the size of the parallax, the registration can still be stable.

Accurately recovering 3D shapes from single images that contain complex backgrounds remains a longstanding and difficult task. Unlike machines, the human visual system can effortlessly filter out background distractions and utilize extensive geometric and semantic knowledge to interpret 3D structures precisely. In contrast, current single-image 3D reconstruction methods often struggle to focus on the target object when confronted with complex backgrounds, as noise and irrelevant objects reduce reconstruction accuracy. To address this issue, we propose a cross-modal fusion strategy that integrates feature enhancement with difference-guided attention to enable high-quality 3D reconstruction from single complex images (named DGGR-Net). Specifically, we utilize retrieved 3D models from the ShapeNet dataset as structural priors and introduce a local geometry-preserving graph convolution module (LGPConv) to optimize fine-grained point cloud structures. Additionally, we design a bidirectional spatial attention (BSA) module to effectively capture spatial image features, reducing background interference during feature extraction. Furthermore, we propose a difference-guided cross-modal attention (DCA) module, which explicitly computes the differences between image and point cloud features to guide precise cross-modal feature fusion, thereby improving modality complementarity and robustness. The experimental results show that our proposed method achieves a Chamfer Distance (CD) of \(3.18 \times 10^{-2}\) and an Earth Mover’s Distance (EMD) of \(3.70 \times 10^{-2}\) on the ShapeNet dataset, and \(5.62 \times 10^{-2}\) and \(7.30 \times 10^{-2}\) respectively on the Pix3D dataset. Compared with the latest method RGB2Point, our method achieved an average improvement of approximately 13.32% and 12.55% in both CD and EMD metrics across the ShapeNet and Pix3D datasets.

Due to the often substantial size of the real-world point cloud data, efficient transmission and storage have become critical concerns. Point cloud compression plays a decisive role in addressing these challenges. Recognizing the importance of capturing global information within point cloud data for effective compression, many existing point cloud compression methods overlook this crucial aspect. To tackle this oversight, we propose an innovative end-to-end point cloud compression method designed to extract both global and local information. Our method includes a novel Transformer module to extract rich features from the point cloud. Utilization of a pooling operation that requires no learnable parameters as a token mixer for computing long-distance dependencies ensures global feature extraction while significantly reducing both computations and parameters. Furthermore, we employ convolutional layers for feature extraction. These layers not only preserve the spatial structure of the point cloud, but also offer the advantage of parameter independence from the input point cloud size, resulting in a substantial reduction in parameters. Our experimental results demonstrate the effectiveness of the proposed TransPCGC network. It achieves average Bjontegaard Delta Rate (BD-Rate) gains of 85.79% and 80.24% compared to Geometry-based Point Cloud Compression (G-PCC). Additionally, in comparison to the Learned-PCGC network, our approach attains an average BD-Rate gain of 18.26% and 13.83%. Moreover, it is accompanied by a 16% reduction in encoding and decoding time, along with a 50% reduction in model size.

In this paper, we present a novel photometric compensation network named CASEformer, which is built upon the Swin module. For the first time, we combine coordinate attention and channel attention mechanisms to extract rich features from input images. Employing a multi-level encoder-decoder architecture with skip connections, we establish multiscale interactions between projection surfaces and projection images, achieving precise inference and compensation. Furthermore, through an attention fusion module, which simultaneously leverages both coordinate and channel information, we enhance the global context of feature maps while preserving enhanced texture coordinate details. The experimental results demonstrate the superior compensation effectiveness of our approach compared to the current state-of-the-art methods. Additionally, we propose a method for multi-surface projection compensation, further enriching our contributions.