改进的灰度-局部方差二维直方图图像分割
发布时间：2018-08-14 09:24
【摘要】：为了提高图像在受到高强度高斯噪声影响下的分割效果,针对传统的二维直方图灰度-平均灰度法,平均灰度-梯度法,二维Otsu斜分法等方法一致性低、对比度低和分割不够准确的情况,现提出一种改进的二维直方图灰度-局部方差的方法。局部方差不仅综合考虑了各像素点与中心像素点数据的离散程度,而且降低了图像受噪声干扰的影响。为了提高分割速度,减少计算量,使用了快速递推算法。实验结果表明该方法比传统的Otsu灰度-平均灰度法和平均灰度-梯度法具有更好的分割效果、一致性和对比度更高。
[Abstract]:In order to improve the segmentation effect of image under the influence of high intensity Gao Si noise, the traditional methods such as gray-average gray scale method, average gray-gradient method, two-dimension Otsu oblique method and so on, have low consistency. In the case of low contrast and inaccurate segmentation, an improved method of gray-local variance of two-dimensional histogram is proposed. The local variance not only considers the discrete degree of each pixel point and the center pixel data, but also reduces the influence of noise interference on the image. In order to improve the speed of segmentation and reduce the amount of computation, a fast recursive algorithm is used. The experimental results show that this method has better segmentation effect and higher consistency and contrast than the traditional Otsu gray-average gray-scale method and average gray-gradient method.
【作者单位】： 湘潭大学物理与光电物理学院;湘潭大学信息工程学院;
【基金】：湖南省自然科学基金资助项目(No.14JJ3077) 湖南省教育厅一般项目(No.13C917)
【分类号】：TP391.41
[Abstract]:In order to improve the segmentation effect of image under the influence of high intensity Gao Si noise, the traditional methods such as gray-average gray scale method, average gray-gradient method, two-dimension Otsu oblique method and so on, have low consistency. In the case of low contrast and inaccurate segmentation, an improved method of gray-local variance of two-dimensional histogram is proposed. The local variance not only considers the discrete degree of each pixel point and the center pixel data, but also reduces the influence of noise interference on the image. In order to improve the speed of segmentation and reduce the amount of computation, a fast recursive algorithm is used. The experimental results show that this method has better segmentation effect and higher consistency and contrast than the traditional Otsu gray-average gray-scale method and average gray-gradient method.
【作者单位】： 湘潭大学物理与光电物理学院;湘潭大学信息工程学院;
【基金】：湖南省自然科学基金资助项目(No.14JJ3077) 湖南省教育厅一般项目(No.13C917)
【分类号】：TP391.41
【相似文献】
相关期刊论文 前10条
1 向新程;丛鹏;陈景运;;二维直方图阈值法分割工业CT图像的研究[J];原子能科学技术;2007年03期
2 吴一全;潘U，
本文编号：2182417
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2182417.html
