液晶与显示, 2013, 28 (3): 417, 网络出版: 2013-07-16  

Bayer图像色彩还原线性插值方法

Linear Interpolation of Demosaicing for Bayer Pattern
作者单位
中国科学院 长春光学精密机械与物理研究所 中国科学院航空光学成像与测量重点实验室,吉林 长春130033
摘要
基于Bayer模板的单CCD 相机通过彩色滤波阵列采集图像数据,图像中的每个像素只有红、绿、蓝3种颜色分量中的一种, 为提高图像分辨率, 获取高质量的彩色图像, 需要通过色彩还原方法恢复另外两种颜色分量。线性插值方法算法简单、运算速度快, 在数字图像色彩还原中经常采用。文章结合Bayer模板结构特点, 介绍目前常用的3种色彩还原插值算法, 包括基于像素复制法、双线性插值法和高质量插值法, 分析和讨论了这些算法的思路和性能特点。通过样本图像色彩还原对比实验, 对各种算法进行了图像质量和处理速度性能比较, 结果表明,高质量插值法还原图像细节丰富, 具有最佳的色彩还原性能。文中最后对色彩还原技术的未来发展趋势进行了展望。
Abstract
Single CCD digital color cameras generally use a Bayer mask over the CCD.Bayer filter mosaic is a color filter array for arranging color filters on a square grid of photosensors and each CCD element captures either red, green or blue light. The acquired images have only one of the three color components at each pixel location. In order to improve quality and resolution of the color image, it is necessary that the Bayer pattern demosiacing algorithm is used to reconstruct the missing two components at each location. The linear interpolation algorithm is fast and easy, that has been applied to the demosiacing for image of digital camera. The three classical linear interpolation algorithms based on Bayer pattern is presented in detail, including a pixel doubling interpolation, a bilinear interpolation and a high quality linear interpolation. The basic theories and performance characteristics of the three methods are analyzed. The comparison of the interpolation accuracy as well as the speed of the different methods is based on the sample images.The results show that the high quality linear interpolation has best results on demosiacing performance. And lastly future development of demosiacing is discussed.

孙辉, 柏旭光, 孙丽娜, 李志强. Bayer图像色彩还原线性插值方法[J]. 液晶与显示, 2013, 28(3): 417. SUN Hui, BAI Xu-guang, SUN Li-na, LI Zhi-qiang. Linear Interpolation of Demosaicing for Bayer Pattern[J]. Chinese Journal of Liquid Crystals and Displays, 2013, 28(3): 417.

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!