摘要
在视频信号处理中,常常需要将采集到的色彩图像转换为灰度图像再做进一步的分析处理.传统的线性转换算法无法保持色彩差异信息,而非线性算法的时间开销太大或需要人工交互,因此应用并不广泛.为解决上述问题,本文提出了一种基于高斯混合模型的实时色彩——灰度转换算法,将色彩按照R,G,B3个分量所占的比例进行分类,保证了在色彩空间中差异较大的颜色也能够分配到差异较大的灰度级别上.同时,我们还加入了动态色彩转换表的机制,使得灰度转换的速度得到了很大的提高.实验结果表明,该算法能够有效地保留图像中的视觉信息,并满足视频实时灰度转换的要求.
In the process of video signal,we usually need to convert color images to grayscale images for further analysis and process.Traditional linearity algorithm can not maintain vision information and nonlinearity algorithm requires high time complexity or manual interaction.Therefore,their application is not widespread.In order to solve the above problems,we present an algorithm for real-time color——grayscale transfer based on Gaussian mixture model.We classify a color into three categories according to the proportional of R,G,B components in the color,thus the color with large difference in the color space can also be distributed on the different gray level.Furthermore,a dynamic color conversion table is incorporated in our algorithm to accelerate transfer process.The experimental results indicate that our algorithm can effectively retain the vision information delivered by the color difference and meet real-time requirements.
出处
《云南大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第S1期31-36,共6页
Journal of Yunnan University(Natural Sciences Edition)
基金
云南大学学术创新团队建设计划
