【24h】

An Effective Integer Demosaicing Based on Directional Filtering

机译:基于方向滤波的有效整数去马赛克

获取原文

摘要

Nowadays many digital still cameras and mobile phones use single image sensor to capture the colors of a scene. In this way, only one of the three primary colors is obtained at each pixel location and the other two are missing. The goal of demosaicing or color interpolation is to estimate the missing colors so that a full-color image can be reconstructed. Although many demosaicing algorithms have been proposed, few of them are suitable for the implementation of integer (or fixed-point) processors and give good performance. To solve this problem, we present an effective integer demosaicing technique based on directional filtering. An appealing property is that all the divisions in the technique can be implemented by using only shifts. Moreover, a new directional refining step is proposed to improve the reconstructed image quality. Experimental results show that the proposed technique can give better performance than several integer techniques.
机译:如今,许多数码相机和移动电话都使用单个图像传感器来捕获场景的颜色。以此方式,在每个像素位置仅获得三种原色中的一种,而其他两种缺失。去马赛克或颜色插值的目的是估计丢失的颜色,以便可以重建全色图像。尽管已提出了许多去马赛克算法,但其中很少有一种适合于整数(或定点)处理器的实现,并且具有良好的性能。为了解决这个问题,我们提出了一种基于方向滤波的有效整数去马赛克技术。一个吸引人的特性是,该技术中的所有划分都可以仅通过使用移位来实现。此外,提出了一种新的方向细化步骤,以改善重建的图像质量。实验结果表明,与几种整数技术相比,该技术具有更好的性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号