基于粒子群优化算法的Kapur熵多阈值图像分割  

Kapur Entropy Based on Particle SwarmOptimization Algorithmfor Multilevel Thresholding Image Segmentation

加关注

下载全文

作  者:马军 鲍日洋 彭晓旭 宋苏航 陈培昕 孙康健 MA Jun;BAO Ri-yang;PENG Xiao-xu;SONG Su-hang;CHEN Pei-xin;SUN Kang-jian(College of Mechanical and Electrical Engineering,Northeast Forestry University,Harbin 150040 China)

机构地区:[1]东北林业大学机电工程学院,黑龙江哈尔滨150040

出  处:《科技创新与生产力》2018年第11期79-81,85共4页Taiyuan Science and Technology

基  金:东北林业大学大学生国家级创新训练计划项目(201810225049).

摘  要:本文将多阈值图像分割中的最佳阈值向量优化问题作为研究对象,采用粒子群优化算法(PSO)对Kapur熵阈值图像分割法的最佳阈值向量进行寻优。传统的Kapur熵阈值图像分割法存在算法运算速度慢、精度不高等问题,本文提出的基于粒子群优化算法的Kapur熵多阈值图像分割法改善了上述问题,具有算法运行快、分割精度高的特点。利用本文提出的方法在阈值等级分别为1,2,3,4,5的情况下进行实验,并应用峰值信噪比(PSNR)指标和结构相似性(SSIM)指标对分割后的图像进行评估。实验结果表明,本文提出的基于粒子群优化算法的Kapur熵多阈值图像分割法优于传统的Kapur熵多阈值图像分割法,可以更加高效地对复杂图像进行多阈值图像分割处理,具有较强的实用性。In this paper, the optimal threshold vector in multilevel thresholding image segmentation is taken as the optimization problem. Particle Swarm Optimization (PSO) is used to optimize the optimal threshold vector of Kapur entropy segmentation method. The traditional Kapur entropy method has the problems of slow algorithm speed and low precision. The Kapur entropy based on particle swarm optimization (PSO) proposed method in this paper improves the above problems, and has the characteristics of fast operation and high segmentation accuracy. The proposed method is tested with threshold levels of 1, 2, 3, 4 and 5 respectively, and the segmented images are evaluated with PSNR (Peak Signal to Noise Ratio) and SSIM (Structural Similarity Index). The experimental results show that the proposed segmentation method is superior to the traditional Kapur entropy method, and can be more efficient in multilevel thresholding segmentation of complex images with strong practicability.

关 键 词:图像分割 多阈值图像分割 粒子群优化算法 Kapur熵 

分 类 号:TP391.41[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象