计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (27): 243-245.

• 工程与应用 • 上一篇    下一篇

改进的遗传算法在白细胞识别中的应用研究

吴建斌1,2,李太全3,田 茂2   

  1. 1.华中师范大学 信息技术系,武汉 430079
    2.武汉大学 电信学院,武汉 430072
    3.长江大学 油气资源与勘探技术教育部重点实验室,湖北 荆州 434002
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-09-21 发布日期:2007-09-21
  • 通讯作者: 吴建斌

Research about improved genetic algorithm applied in white cell’s auto recognition

WU Jian-bin1,2,LI Tai-quan3,TIAN Mao2   

  1. 1.Department of Information Technologic,Huazhong Normal University,Wuhan 430079,China
    2.School of Electronic Information,Wuhan University,Wuhan 430072,China
    3.Key Laboratory of Exploration Technologies for Oil and Gas Resources,Yangtze University,Jinzhou,Hubei 434002,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-09-21 Published:2007-09-21
  • Contact: WU Jian-bin

摘要: 针对多光谱图像数据维数高,数据量大的特点,鉴于自适应遗传算法在搜索最优解上特有的优点,提出了采用自适应遗传算法进行白细胞的特征提取,同时为了增强算法的稳定性,提高收敛速度,部分改进了原算法。在此基础上,利用选取的53个特征和二值支持向量机相结合,构造分类器,有效地解决了白细胞的分类识别问题。实验结果表明,改进后的算法具有更快的收敛速度,更好的稳定性,设计的分类器有效地提高了识别速度和精度,识别率达89.02%。

关键词: 特征选择, 遗传算法, 支持向量机, 模式识别

Abstract:

For the high-dimension data of multi-spectral microscope cell image,a white blood cell’s feature selection method based on Adaptive Genetic Algorithm(AGA) is presented in this paper.The paper improves the AGA for promoting convergence speed and high robustness.Based on selecting 53 white blood cell’s features,the paper applies the Support Vector Machines(SVM) to realize the white blood cell recognition.The result approves the way of feature selection and classification recognition is valid,the ration recognition can reach 89.02%.

Key words: feature selection, Genetic Algorithm, Support Vector Machines(SVM), pattern recognition