基于交叉皮质模型的高动态范围图像可视化方法  

Visualization of High Dynamic Range Image based on Intersecting Cortical Model

作　　者：李成[1] 王晨[1] 鞠明[1] 高山[1] 毕笃彦[1] 

机构地区：[1]空军工程大学工程学院,西安710038

出　　处：《计算机辅助设计与图形学学报》2011年第8期1420-1426,共7页Journal of Computer-Aided Design & Computer Graphics

基　　金：航空科学基金(20101996009)

摘　　要：针对高动态范围图像(HDRI)在低动态范围显示设备上的可视化问题,提出基于交叉皮质模型(ICM)的色调映射算法,其在融入人类视觉特性的同时兼顾了全局亮度和局部细节保持.首先将线性热流(LHF)应用于向心自动波(CA)的实现,构成LHF-CA-ICM,以解决原有ICM自动波效应的负面影响;然后根据LFH-CA-ICM迭代过程,通过设计自适应S形函数将亮度信息从高动态范围向低动态范围映射,并进行基于直方图截取的后处理;最后在RGB空间恢复颜色信息.实验结果表明,该算法可以获得HDRI高视觉质量显示效果,能够再现真实场景的细节和丰富色彩,并有效地避免了光晕等人工效应.In this paper, we propose an algorithm based on intersecting cortical model （ICM） to visualize the high dynamic range images （HDRI） on low dynamic range device. Our algorithm focuses on human visual characteristics and the combination of both global lightness and local detail. Based on linear heat flow theory （LHF）, the centripetal autowave （CA） is used to form the LHF-CA-ICM. It solves the interference from ICM＇s autowave effect, which could blur the edge and detail. During the iterations of LHF-CA-ICM, we propose an S-shaped function as the adaptive tone mapping operator （TMO）. Black and white point correction and nonlinear color restoration in RGB color space are conducted for better display performance. Experimental results show that HDRIs obtain good visualization with much more realistic detail and color, and the halo artifact is reduced.

关 键 词：交叉皮质模型 高动态范围图像 色调映射算子 映射函数 向心自动波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]