研究目的
To propose a method to decrease temporal artifacts in tone mapped HDR video sequences, specifically addressing noise, flicker, and temporal inconsistencies.
研究成果
The temporal adaptation method effectively reduces noise, flicker, and temporal inconsistencies in HDR video tone mapping, providing visual comfort and stability. It shows promise for real-time applications and future work in areas like video assistance and autonomous systems.
研究不足
The method may not fully address all types of temporal artifacts and is tested only in specific indoor scenarios; optimization for different lighting conditions and real-time performance on various hardware could be explored.
1：Experimental Design and Method Selection:
The method involves temporal filtering of the maximum luminance value in each frame before applying a tone mapping function, inspired by Drago et al.'s adaptive logarithmic tone mapping.
2：Sample Selection and Data Sources:
Videos were captured using a logarithmic sensor camera (NIT MC1003-1VC) in indoor environments, including sequences with no lighting variation, small blinking LED source, and abrupt variations from a halogen lamp.
3：List of Experimental Equipment and Materials:
Logarithmic sensor camera (NIT MC1003-1VC), I5-4200U processor for processing.
4：Experimental Procedures and Operational Workflow:
Videos are processed using both the original Drago algorithm (eq. 3) and the proposed temporal adaptation (eq. 6), with parameters Ldmax = 100 cd/m2 and b =
5：85, and temporal filtering over 17 frames. Data Analysis Methods:
Results are evaluated based on luma level average, range level evolution, and visual inspection for artifacts like flicker and temporal incoherency.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
