计算机工程与应用 ›› 2013, Vol. 49 ›› Issue (23): 127-131.

• 图形图像处理 • 上一篇    下一篇

基于改进自适应降雪模型的随机游走图像分割

朱  强   

  1. 浙江传媒学院 教育技术中心,杭州 310018
  • 出版日期:2013-12-01 发布日期:2016-06-12

Random walking image segmentation based on improved adaptive snowfall model

ZHU Qiang   

  1. Educational Technology Center, Zhejiang University of Media and Communications, Hangzhou 310018, China
  • Online:2013-12-01 Published:2016-06-12

摘要: 为了提高图像分割的准确度,尽可能降低分割边缘噪声对图像分割的影响,提出了一种基于降雪模型的图像分割方法。对降雪模型及积雪表面效应做了详细分析,得出降雪模型运用于图像分割具有较强的适应性;接着在传统的随机游走图像分割算法中加入了自适应降雪模型的特性,生成新的算法;运用虚拟图像和真实图像进行算法性能实例仿真,结果表明,该算法的图像分割性能优于常见的NCut和传统随机游走图像分割算法,具有一定的研究价值。

关键词: 图像分割, 降雪模型, 随机游走, 高斯核函数

Abstract: In order to improve accuracy of image segmentation, reduce the effect of noise on the cutting edge of image segmentation as much as possible, a new image segmentation method based on the model of the snowfall is proposed. Snowfall model and snow surface effect are analyzed in detail, the snow model is applied to image segmentation with strong adaptability, and it mixs the traditional random walk image segmentation algorithm with adaptive snow model characteristics, generates a new algorithm, makes performance simulation using virtual and real images algorithm, the results show that the image segmentation performance is better than the common NCut and the traditional random walk algorithm for image segmentation, and it has certain research value.

Key words: image segmentation, snowfall model, random walk, Gauss kernel function