...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Computational Intelligence and Electronic Systems >Demosaicing and Sharpening of Color Filter Captured Images Using Adaptive Transformation
【24h】

Demosaicing and Sharpening of Color Filter Captured Images Using Adaptive Transformation

机译：使用自适应变换对彩色滤光片捕获的图像进行去马赛克和锐化

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Obtaining cost-effective images with diagnosis quality is a current challenge, and this paper proposes a noise removal algorithm for color images corrupted by additive Gaussian noise and a robust open close sequence filter based on mathematical morphology for high probability additive Gaussian noise removal is given. First, an additive Gaussian noise detector using mathematical residues is to identify pixels that are contaminated by the additive Gaussian noise. Then the image is restored using specialized open-close sequence algorithms that apply only to the noisy pixels. But the color blocks that degrade the quality of the image will be recovered by a block smart erase method. This algorithm can be applied to highly corrupted images. Mathematical morphology is a nonlinear image processing methodology that is based on the application of lattice theory to spatial structures. In color images, algorithms are developed for boundary extraction via a morphological gradient operation and for region partitioning based on texture content. Mathematical morphological operations are useful in smoothing and sharpening, which often are useful as per or post processing steps.
机译：获得具有诊断质量的具有成本效益的图像是当前的挑战，本文提出了一种针对加性高斯噪声破坏的彩色图像的噪声去除算法，并给出了一种基于数学形态学的鲁棒开闭序列滤波器，用于高概率性加性高斯噪声去除。首先，使用数学残差的加性高斯噪声检测器将识别被加性高斯噪声污染的像素。然后使用仅适用于噪点像素的专用开闭序列算法还原图像。但是，会通过块智能擦除方法恢复降低图像质量的色块。该算法可以应用于高度损坏的图像。数学形态学是一种非线性的图像处理方法，该方法基于将晶格理论应用于空间结构。在彩色图像中，开发了用于通过形态学梯度运算进行边界提取以及基于纹理内容进行区域划分的算法。数学形态学运算可用于平滑和锐化，这通常在每个或每个后处理步骤中都有用。

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