The evolution of 3D multi-media technology has spurred the need for more effective 3D video storage and transmission methods. Most of the current standardizing 3D volumetric video coding methods in the mesh decimation stage are still limited: the difference of geometric structure in each simplified frame hinders optimal compression. The proposed Tracked QEM Algorithm effectively tracks mesh models across successive frames, offering a tailored solution for dynamic meshes in 3D volumetric videos.

The Tracked QEM Algorithm ensures that the simplified results have better topological consistency and spatial smoothness between consecutive frames than the original QEM algorithm. The evaluation results based on temporal consistency show that the proposed approach is superior to the conventional mesh simplification. The smoother simplified results with similar topology delineate the discontinuous structural information between frames. As a novel pre-processing approach to 3D video compression, this proposal has the potential to improve the compression rate.

Most state-of-the-art methods for pixel-wise hyperspectral image (HSI) classification are based on the Convolutional Neural Network (CNN). In this paper, we introduce a feature reconstruction module (FRM) into the CNN-based network of pixel-wise HSI classification to improve classification accuracy. FRM can extract essential characteristics in the original matrix of CNN features by low-rank approximation using matrix factorization. We compare the classification accuracy before and after the introduction of FRM into the CNN-based network of pixel-wise HSI classification to validate its effectiveness. Experimental results demonstrate this method improved classification accuracy. We also visualized and compared the original CNN features and the reconstructed CNN features to evaluate which features contributed to the improvement in classification accuracy.