对于包含运动物体的复杂场景而言,使用计算机生成用于三维显示的超多视点图的效率往往较低.为解决该问题,本文提出了一种基于反向光线跟踪技术和感兴趣区域(region of interest,ROI)的生成加速算法,在保证显示质量的前提下对超多视点图...
详细信息
对于包含运动物体的复杂场景而言,使用计算机生成用于三维显示的超多视点图的效率往往较低.为解决该问题,本文提出了一种基于反向光线跟踪技术和感兴趣区域(region of interest,ROI)的生成加速算法,在保证显示质量的前提下对超多视点图的生成过程进行加速.在每一帧中,内容变化的部分被检测并标记为ROI,只有该部分的像素被重新渲染.实验结果表明,内容生成过程的速率可被提升至原来的2~9倍.此外,在分辨率为3840×2160的复杂应用情况下,输出帧率由3.5帧/秒提升至超过30帧/秒,满足了实时交互的需求.
This paper presents a wearable piezoelectric generator in order to explore the possibility of harvesting energy from human motion to supply power for wearable or portable electronics, thus acting as alternatives to th...
详细信息
This paper presents a wearable piezoelectric generator in order to explore the possibility of harvesting energy from human motion to supply power for wearable or portable electronics, thus acting as alternatives to the use of *** generator is characterized by a high durability and portability, which is thin enough to be placed in a shoe or a bag *** prototypes has been built and tested both by a material testing system (MTS) and by human motion during walking at 1 *** prototype enables a mean output power of 1 mW, namely 1 mJ per *** seems that the generator cannot provide sufficient energy to run electronics continuously, therefore two methods are proposed to make the generator compatible with *** work demonstrates the feasibility of harvesting the mechanical energy dissipated in human motion for powering low-power electronics.
暂无评论