An infrared image segmentation method based on within-class absolute difference and chaotic particle swarm optimization

被引：0

作者：

Wu Y. ^{[1
]}

Zhan B. ^{[1
]}

Wu J. ^{[1
]}

机构：

[1] School of Information Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing

来源：

Guangxue Xuebao/Acta Optica Sinica | 2010年 / 30卷 / 01期

关键词：

Area difference between background and target; Chaotic mutation; Image processing; Infrared image segmentation; Niche particle swarm optimization; Within-class absolute difference;

D O I：

10.3788/AOS20103001.0079

中图分类号：

学科分类号：

摘要：

A thresholding method for infrared target image is proposed, which is based on the within-class absolute difference, the area difference between background and target, and Niche chaotic mutation particle swarm optimization (NCPSO). The less within-class absolute difference can make the cohesion performance better, and the area difference between background and target is used to inhibit the tendency of an equal division. Therefore, a more reasonable threshold selection rule is formed comprehensively. First, one-dimensional threshold selection method is proposed. The anti-noise performance is improved obviously by extending to the two-dimensional histogram from one-dimensional method. Then the computational burden of finding optimal threshold vector is large for the two-dimensional thresholding, thus NCPSO is used to find the optimal threshold vector. Finally, the proposed method is compared with Fisher method, the Otsu method and the maximum entropy method. The experimental results show that the proposed method is effective for less target infrared image thresholding and the running time is significantly reduced.

引用

页码：79 / 85

页数：6

共 24 条

[1]

Guan Z., Chen Q., Qian W., Et al., An adaptive background adjusting algorithm for dim target detection, Acta Optica Sinica, 27, 12, pp. 2163-2168, (2007)

[2]

Yao Y., Li Y., Liu B., Et al., Analysis of real photonic crystal fibers by finite-difference frequency-domain method combined with digital image processing, Acta Optica Sinica, 28, 7, pp. 1384-1389, (2008)

[3]

Sun Z., Peng Z., Liu H., Et al., Calculated methods of high power laser near field parameters, Chinese J. Lasers, 35, 4, pp. 544-548, (2008)

[4]

Yuan J., Liu R., Gao X., Et al., Preprocessing technology of retina optical coherence tomography image, Acta Optica Sinica, 27, 3, pp. 419-423, (2007)

[5]

Nie S., Wang M., Liu F., Image segmentation algorithm study for low contrast image, Chinese J. Lasers, 31, 1, pp. 89-91, (2004)

[6]

Guo H., Tian T., Wang L., Et al., Image segmentation using the maximum entropy of the two-dimensional bound histogram, Acta Optica Sinica, 26, 4, pp. 506-509, (2006)

[7]

Zhou X., Liu Y., A study of automatic extraction method of defects on X image, Acta Optica Sinica, 26, 7, pp. 1016-1020, (2006)

[8]

Wu Y., Zhu Z., 30 Years(1962-1992) of the developments in threshold selection methods in image processing(1), J. Data Acquisition & Processing, 8, 3, pp. 193-201, (1993)

[9]

Wu Y., Zhu Z., 30 Years(1962-1992) of the developments in threshold selection methods in image processing(2), J. Data Acquisition & Processing, 8, 4, pp. 268-282, (1993)

[10]

Otsu N., A threshold selection method from gray-level histogram, IEEE Trans. Syst. Man and Cybernetics, 9, 1, pp. 62-66, (1979)

← 1 2 3 →