浙江农业学报

• 生物系统工程 • 上一篇    下一篇

基于全变分去噪和八方向sobel算子的叶脉提取算法

  

  1. (沈阳农业大学 信息与电气工程学院, 辽宁 沈阳 110161)
  • 出版日期:2015-04-25 发布日期:2015-04-30

The vein extraction algorithm based on total variation denoising and eight\|direction sobel operator

  1. (College of Information and Electrical Engineering,Shenyang Agricultural University,Shenyang 110161, China)
  • Online:2015-04-25 Published:2015-04-30

摘要: 分析了已有的sobel边缘检测算子在植物叶脉提取不准确的原因。针对在主干叶脉及一些细小叶脉处理效果的不足,提出了一种新的改进算法。该算法首先结合全变分(TV)去噪模型对叶片进行去噪处理,然后基于叶脉走向的多方向性特点,在已有的2个或4个方向模板的基础上,增加其他方向的模板,构造新的包含8个方向的模板算子,利用此模板算子实现植物叶片的叶脉提取。试验结果表明,提出的算法与传统的叶脉提取算法比较,既能较好地去除噪声,又能保留叶脉的细节信息,达到充分提取主干及细小的叶脉的效果。

关键词: sobel边缘检测, 八方向模板, TV去噪

Abstract: The inaccuracy of the existing sobel edge detection operator in extraction of plant veins was analyzed. For lack of treatment effect on veins in the trunk and some small veins, a new improved algorithm was proposed. Combining with the total variation (TV) denoising model, this algorithm method dealt with the denoising of leaves. Then, based on the multi direction characteristics of vein trend, in two or four directions of existing template, the other direction template was considered, a new one containing eight direction template operator was constructed. And this template operator was used for veins extraction of plant leaves. Experimental results showed that compared with the traditional algorithm, the proposed algorithm can not only remove the noise effectively, but also keep the detail information of vein trunk, which can fully extract the trunk and small veins.

Key words: sobel edge detection, eight\, direction template, TV denoising