高动态范围成像已经成为数字显微镜发展的一大趋势。目前视频级高动态范围显微成像技术由于计算复杂度较高, 还无法做到实时处理和显示。提出利用高清电视领域中的 REC.2020色域变换, 结合 GPU实时加速处理技术, 成功实现了高动态范围显微图像视频流的实时显示, 图像具有良好的色彩还原性和细节恢复性。实验结果表明, 设计的算法不仅能够提供稳定的高清、高动态范围视频, 其细节和色彩还原性也有了很大的提高, 突破了传统的高动态范围图像技术在成像速度和质量方面的限制。

High dynamic range (HDR) technology has become a new trend of the development of digital microscopes. Currently, due to the high computational complexity, the HDR video cannot be processed and displayed in real-time. Based on REC.2020 high definition (HD) standard, it is feasible and stable to generate the HDR video flow in real-time using GPU accelerated processing. The experimental results show that the presented algorithm not only provides a stable real-time HDR video flow, but also improves the performance of the recovered video in both detail and color, which has overcome the limitation in processing speed and quality of the traditional HDR imaging technology.