基于改进的NCSPSO优化SVM的木材缺陷图像分割

Wood Defect Image Segmentation Based on Improved NCSPSO's Optimization of SVM

针对木材缺陷图像,采用基于小生境和交叉算子的粒子群算法(NCSPSO)优化支持向量机(SVM)进行图像分割,提取木材缺陷分割图。主要对NCSPSO算法进行改进,寻找最优惩罚系数C和高斯核函数中的参数,然后采用SVM方法对训练样本进行综合训练,以建立最佳分类模型,并对木材缺陷图像分割测试。与模拟退火法(SA)及原NCSPSO算法进行对比实验,从而验证改进NCSPSO算法的优越性。

In view of wood defect images,Particle Swarm Optimization（NCSPSO）algorithm based on microhabitat and crossover operator are adopted to optimize support vector machine（SVM）for image segmentation and wood defect segmentation diagram extraction. The focus is placed on the improvement of NCSPSO to seek the optimal penalty coefficient C and the parameter of Gauss kernel function. Then SVM is used to establish the best classification model and test wood defect image segmentation. Simulate anneal arithmetic and the NCSPSO are compared through experiments to verity the superiority of NCSPSO.