| 基于Ostu优化的PCNN电力故障区域提取 |
| |
| 引用本文： | 林亚君、陈学军.基于Ostu优化的PCNN电力故障区域提取[J].福建工程学院学报,2020,0(6):593-597. |
| |
| 作者姓名： | 林亚君、陈学军 |
| |
| 作者单位： | 莆田学院 |
| |
| 摘 要： | 为更好实现红外图像中电力设备故障区域的提取，提出改进PCNN(脉冲耦合神经网络) 的故障区域提取方法。基于Otsu 算法(大津法，又称最大类间差法) 计算红外图像的最优分割阈值，作为PCNN 迭代的初始阈值。以最大类间方差作为PCNN 模型的收敛判据，实现红外图像自动分割以提取电力故障区域。实验表明，该算法与Otsu、K-means、传统PCNN 方法相比，能够更全面、精确地提取电力故障区域，为后续故障类型识别奠定基础。
|
| 关 键 词： | 红外图像 大津法 脉冲耦合神经网络 图像分割 电力故障 |
Fault zone extraction of electrical equipment by using PCNN based on Otsu optimization |
| |
| LIN Yajun,,CHEN Xuejun.Fault zone extraction of electrical equipment by using PCNN based on Otsu optimization[J].Journal of Fujian University of Technology,2020,0(6):593-597. |
| |
| Authors: | LIN Yajun CHEN Xuejun |
| |
| Affiliation: | School of Electromechanical Engineering,Putian University |
| |
| Abstract: | In order to better realize the extraction of the fault area of power equipment in infrared images, a method for fault area extraction was proposed with improved PCNN (pulse-coupled neural network). Based on Otsu, the optimal segmentation threshold for infrared images was calculated based on Otsu algorithm and it was taken as the initial threshold for PCNN iterations. By using the maximum between-cluster variance as the convergence criterion of the PCNN model, the infrared image was automatically segmented to extract the power fault region. Experimental results show that compared with Otsu, K-means and traditional PCNN methods, this algorithm can extract the power fault area more comprehensively and accurately, and lay the foundation for the identification of subsequent fault types. |
| |
| Keywords: | infrared image Otsu PCNN image segmentation electrical equipment fault |
|
| 点击此处可从《福建工程学院学报》浏览原始摘要信息 |
|
点击此处可从《福建工程学院学报》下载全文 |
