Compression and Rendering of Space Environment Volume Data Based on Improved HVQ Algorithm

Abstract

We propose an improved HVQ approach by combining two detail levels to compress space environment volumes, based on the variation characteristics including smooth variation and significantly positive correlation between variations in the same direction at two different scales. First, the space environment data is divided into 43 blocks. Then blocks are decomposed into a two-level hierarchical representation and each block is represented by a mean value and a detail vector. Finally, the detail vectors are encoded by a vector quantizer. During the vector quantization process, PCA-split is applied to compute an initial codebook, and then LBG-algorithm is conducted for codebook refinement and quantization. We take advantage of the codebook-retraining method to speed up the quantization of time series with temporal coherence. Furthermore, we employ the progressive rendering based on GPU for realtime interactive visualization. The results of our experiments prove that the algorithm proposed in this paper can significantly improve the compression rate and decoding speed without sacrificing fidelity in space environment domain.

Design and Implementation of Augmented Image for the Space Environment Journals Based on AR Technology

“…Finally we simulate 3D space environment scene, display in a virtual and real fusion manner, and respond to user actions. In order to overcome the difficulties in target image database update, space environment image recognition and massive time-varying 3D data rendering, we adopt and improve the web framework architecture of Operational Space Environment Technology System (OSETS) [4], SIFT-TF-ORB algorithm for image recognition [8], and HVQ-1d method for compression and rendering of space environment volumes [9] in system design.…”

“…This makes the rendering and fusion process difficult and sometimes even impossible because of the limited memory resources. Bao et al [9] proposed HVQ-1d, a compressed volume rendering method, to pre-process and render space environment volume data. HVQ-1d can significantly decrease data size without scarifying fidelity.…”

“…Existing AR applications for journals adopt static target image database containing images with many unique features, and render simple 3D objects. While in space environment, most images contain similar patterns and gradient distributions [8], target image database needs to be updated periodically, and models always generate massive time-varying 3D data sets [9,10]. These characteristics and demands make AR applications difficult to realize.…”