arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1009.0854cs.CV
收藏

基于极大极小近似的快速颜色空间变换

Fast Color Space Transformations Using Minimax Approximations

Georgia Institute of Technology(佐治亚理工学院);Wayne State University(韦恩州立大学)

M. Emre Celebi, Hassan Kingravi, Fatih Celiker

摘要:

颜色空间变换经常用于图像处理、图形和可视化应用中。在许多情况下,这些变换是复杂的非线性函数,这禁止它们在时间关键的应用中使用。本文提出了一种新的颜色空间变换的极大极小近似方法(MACT),并在三种常用的颜色空间变换上证明了MACT。大量的实验和广泛的不同的图像集和著名的多维查找表插值方法的比较表明,MACT实现了四个标准之间的良好平衡:易于实现,内存使用,准确性和计算速度。

英文摘要:

Color space transformations are frequently used in image processing, graphics, and visualization applications. In many cases, these transformations are complex nonlinear functions, which prohibits their use in time-critical applications. In this paper, we present a new approach called Minimax Approximations for Color-space Transformations (MACT).We demonstrate MACT on three commonly used color space transformations. Extensive experiments on a large and diverse image set and comparisons with well-known multidimensional lookup table interpolation methods show that MACT achieves an excellent balance among four criteria: ease of implementation, memory usage, accuracy, and computational speed.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑