Improving the Performance of Leaves Identification by Features Selection with Genetic Algorithms
The development of vision systems to plant leaves identification from images is a very important current challenge. One of the main lines of research is the improvement in performance. The performance of automatic identification systems is directly linked to the extracted features. The extracted features allow to identify or classify the object. However, sometimes a high number of features introduces noise which affects the performance. In this research, a genetic algorithm is proposed to extract the most discriminative features. The proposed technique reduces the dimensionality of the data set and improves the performance to identifying plants. In the experimental results, the proposed method is compared with feature selection classic techniques. Experimental results show that the proposed technique obtains a significant improvement in the performance in comparison with other techniques.
KeywordsGenetic Algorithm Feature Selection Textural Feature Classifier Performance Binary String
This research was funded by the Ministry of Research of the Autonomous University of the State of Mexico with the research project 3771/2014/CIB.
- 3.Ma, W., Manjunath, B.: Texture features and learning similarity. In: Proceedings of the Conference on Computer Vision and Pattern Recognition, pp. 425–430 (1996)Google Scholar
- 5.Kohavi, R.: A study of cross-validation and bootstrap for accuracy estimation and model selection. In: Proceedings of the Fourteenth International Joint Conference on Artificial Intelligence, vol. 2(12), pp. 1137–1143. Morgan Kaufmann, San Mateo (1995)Google Scholar
- 28.Mohideen, K.A., Saravanakumar, G., Valarmathi, K., Devaraj, D., Radhakrishnan, T.K.: Real-coded Genetic Algorithm for system identification and tuning of a modified Model Reference Adaptive Controller for a hybrid tank system. Appl. Math. Model. 37(6), 3829–3847 (2013)MathSciNetCrossRefGoogle Scholar