摘要
针对传统图像分割方法中出现的分割精度低以及算法适用性欠佳等问题,该文提出了基于分通道自适应理论的阈值分割算法,并对分割结果进行对比评价。结合Welsh算法中的自适应理论对阈值分割法进行优化,将其运用到10组不同类型的图像分割当中,均取得良好的分割效果。对比7种传统分割算法与该文算法的图像分割结果,该文算法的均匀性测度评价值均大于0.95,准确率、召回率与轮廓精度的值均大于98%,4个评价指标值均接近于1且高于传统的分割算法。结果表明:该文算法使得分割精度得到了提升,且具有更优的适用性。
In order to solve the problems of low precision and poor applicability of traditional image segmentation methods,this paper proposes a threshold segmentation algorithm based on adaptive theory of sub channels and compares and evaluates the segmentation results.Combining with the adaptive theory of Welsh algorithm,the threshold segmentation method is optimized and applied to 10 groups of different types of image segmentation,and good segmentation results are achieved.By comparing the image segmentation results of the 7 traditional segmentation algorithms and the proposed algorithm,the uniformity of intraregion evaluation value of the proposed algorithm is greater than 0.95,the values of accuracy,recall and boundary F-measure are greater than 98%,and the four evaluation index values are close to 1 and higher than the traditional segmentation algorithm.The results show that the proposed algorithm improves the segmentation accuracy and has better applicability.
作者
李洁
于艺铭
陈茜
王小菊
王琪
Li Jie;Yu Yiming;Chen Xi;Wang Xiaoju;Wang Qi(College of Light Industry and Food Engineering,Nanjing Forestry University,Nanjing 210037,China)
出处
《南京理工大学学报》
CAS
CSCD
北大核心
2022年第5期606-614,631,共10页
Journal of Nanjing University of Science and Technology
基金
国家自然科学基金(31870565)。
关键词
图像分割
Lab分通道
自适应理论
阈值
数学形态学理论
image segmentation
Lab sub channels
adaptive theory
thresholds
mathematical morphology theory