Shape feature selection and weed recognition based on image processing and ant colony optimization

被引：0
|
作者：
Li X. [1 ,2 ]
Zhu W. [1 ]
Ji B. [1 ]
Liu B. [1 ]
Ma C. [1 ]
机构：
[1] School of Electronic and Information Engineering, Jiangsu University
[2] School of Information Engineering, Yancheng Institute of Technology
来源：
Nongye Gongcheng Xuebao/Transactions of the Chinese Society of Agricultural Engineering | 2010年 / 26卷 / 10期
关键词：
Ant colony optimization algorithm; Feature extraction; Image processing; Support vector machine;
D O I：
10.3969/j.issn.1002-6819.2010.10.030
中图分类号：
学科分类号：
摘要：
Using shape features of plant leaf to identify the weed from the crop is an important method for weed recognition. In order to improve the accuracy and efficiency, overlapped leaves were separated through morphology operation and threshold segmentation based on distance transformation, and 17 shape features including geometric features and moment invariant features were extracted from the single leaf. Finally, ant colony optimization (ACO) algorithm and support vector machine (SVM) classifier were used to feature selection and weed recognition so as to select more superior features and optimize features combination. The recognition experiment for weed in cotton field demonstrates that this method can reduce the number of features effectively and the accuracy is over 95% by using optimized feature subset. The results show that the proposed method has good performance on accuracy and stability, and can be reference for quick and accurate identifying of weeds.
引用
收藏
页码：178 / 182
页数：4
相关论文
共 14 条
  • [1] Burks T.F., Shearer S.A., Heath J.R., Evaluation of neural-network classifiers for weed species discrimination, Biosystems Engineering, 91, 3, pp. 293-304, (2005)
  • [2] Mao W., Cao J., Jiang H., Et al., In-field weed detection method based on multi-features, Transactions of the CSAE, 23, 11, pp. 206-209, (2007)
  • [3] Sogaard H.T., Heisel T., Weed classification by active shape models, Biosystems Engineering, 91, 3, pp. 271-281, (2005)
  • [4] Schuster I., Nordmeyer H., Rath T., Comparison of vision based and manual weed mapping in sugar beet, Biosystems Engineering, 98, 1, pp. 17-25, (2007)
  • [5] Long M., He D., Weed identification from corn seedling based on computer vision, Transactions of the CSAE, 23, 7, pp. 139-144, (2007)
  • [6] Dorigo M., Blum C., Ant colony optimization theory: A survey, Theoretical Computer Science, 34, 4, pp. 243-278, (2005)
  • [7] Chen F., Hu P., Extracting and identifying point symbols based on distance transform, Geomatics and Information Science of Wuhan University, 32, 12, pp. 1172-1175, (2007)
  • [8] You Y., Fan Y., Pang Q., Separate algorism for overlapping cell images based on distance transformation, Computer Engineering and Applications, 20, 3, pp. 206-208, (2005)
  • [9] Liu X., Xiang J., Yang B., A labeling algorithm based on 8-connected boundary tracking, Computer Engineering and Applications, 23, 2, pp. 125-126, (2001)
  • [10] Hu M.K., Visual pattern recognition by moment invariants, IRE Transaction Information Theory, 8, 2, pp. 179-187, (1962)
12