Image Segmentation Method Based on Fisher Criterion and Genetic Algorithm

Abstract:

Precise recognition of the weed by computer vision, furthermore raising the weeding efficiency, reducing the use of herbicide, and decreasing the pollution to the environment is one of the key technologies in the field of precision agriculture. To determine the optimal threshold in image automatic segmentation and solve one-dimensional histogram without obvious peak and valley distribution, image segmentation method based on fisher criterion and improved adaptive genetic algorithm is proposed. This method can preserve the multifamily of population and the astringency of the algorithm, and can overcome the problems of poor astringency and premature occurrence. The result shows that the proposed approach has better immunity to Salt and Pepper Noise and greatly shortens the time of image segmentation.

• Periodical: Key Engineering Materials (Volumes 474-476)
• Pages: 928-932
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.474-476.928
• Online since: April 2011
• Authors: Xian Xiang Fu, Zu Jue Chen, Yong Fu Zhao
• Keywords: 2-D Histogram, Adaptive Genetic Algorithm (AGA), Fisher Criterion, Image Segmentation

References

[1] W.X. Zhu, F.J. Jin, R.R. Tan: Weed recognition method based on color and morphological features in wheat field. TransactionsComputerApplication. Vol: 27-11(2007), p.2870.

[2] N, A Otsu: Thresholding selection method from gray-level histograms. IEEE Trans System Man Cybernetic. Vol: 21(1979). p.62.

[3] Yu Jinhua, WangYuanyuan: Image segmentation with two-dimension fuzzy cluster method based on spatial information. Opto-Electronic Engineering. Vol: 34-4 (2007), p.114.

[4] J.L. Fan,F. Zhao. Two-Dimensional Otsu's Curve Thresholding Segmentation Method for Gray-Level Images. Acta Electronica sinica. Vol: 35-4 (2007), p.751.

[5] Y.Q. Wu, Z. Pan, W.Y. Wu. Image thresholding based on two-dimensional histogram oblique segmentation and its fast recurring algorithm. Journal on Communication Vol: 29-4(2008). p.77.

[6] Y.Q. Wu, J.K. Zhang: Thresholding Based on Maximum Entropic Correlation of Average Gray Level-gradient Two-dimensional Histogram. Journal of Chinese Computer Systems. Vol: 30-8(2009). p.1675.

[7] J.L. Fan, B. Lei: Fast two-dimensional otsu algorithm based on integral image. Chinese Journal of Scientific Instnlment. Vol: 30-1(2009). p.39.

[8] Wan, Yi; Wang, Jiangchang: A modified Otsu image segment method based on the Rayleigh distribution. ICCSIT Vol: 5(2010). p.281.

DOI: 10.1109/iccsit.2010.5563957

[9] Qingzhen, Xu1; Jing, Kong: A new segmentation algorithm with the fisher criterion function. CCCM Vol: 4(2009), p.63.