摘要

Next generation mobile network has the capacity to realize virtual reality because it can provide low latency, high data rate and low error rate. In virtual reality, instant rendering needs more computing power, but 5G terminal device computing power is relatively weak. Illumination precomputed technology, such as Spherical Harmonic Lighting, provides a mean for rendering scene with limited computing power. For the restrictions on efficiency and storage, Spherical Harmonic Lighting is preferred to capture low-frequency illumination information but missing on high-frequency illumination information. The precomputed illumination dataset will expand rapidly if more high-frequency information is preserved, which prevents its transmission on 5G network. In this paper, we have proposed a method based on wavelet decomposition, that can improve Spherical Harmonic Lighting sampling efficiency and capture more frequency illumination information with adaptive Spherical Harmonic band index. This method can preserve more frequency illumination with relatively small precomputed dataset..

全文