Study on the effect of training samples on the accuracy of crop remote sensing classification

Pan Hongtao; Wang Xuan; Wang Xiaofei

doi:10.3788/IRLA201746.S126003

In order to study and analyse the influence the number of and quality of the training samples on the classification accuracy better, Helen city in Heilongjiang Province was chosen as the research required experimentation area, using Landsat 8 remote sensing images as the data source, the effects of the number and quantity of training samples on the classification accuracy were studied respectively by using the maximum likelihood, neural network and support vector machine three kinds of methods, and several experiments were made on these three kinds of classification methods. The final result shows that:(1) when the training sample quality is relatively constant, the degree of response of the same classification method to the same number of training samples as well as the degree of response of the different classification methods to the number of training samples are different, and the classification accuracy has different degree of volatility, with the increase of the number of training samples, the volatility will decrease, when the number of training samples reaches a certain degree, the mean of classification accuracy will tend to be relatively stable; (2) when the number of training samples is constant, the same classification methods as well as the different classification methods have different degree of response to the training samples of the same quality grade; the degree of response of the same classification method to the different training samples quality level is also different.

Study on the effect of training samples on the accuracy of crop remote sensing classification

1. School of Electrical Engineering,Heilongjiang University,Harbin 150080,China

Received Date: 2017-06-05

Rev Recd Date: 2017-07-09

Publish Date: 2017-12-31

Key words:

Abstract: In order to study and analyse the influence the number of and quality of the training samples on the classification accuracy better, Helen city in Heilongjiang Province was chosen as the research required experimentation area, using Landsat 8 remote sensing images as the data source, the effects of the number and quantity of training samples on the classification accuracy were studied respectively by using the maximum likelihood, neural network and support vector machine three kinds of methods, and several experiments were made on these three kinds of classification methods. The final result shows that:(1) when the training sample quality is relatively constant, the degree of response of the same classification method to the same number of training samples as well as the degree of response of the different classification methods to the number of training samples are different, and the classification accuracy has different degree of volatility, with the increase of the number of training samples, the volatility will decrease, when the number of training samples reaches a certain degree, the mean of classification accuracy will tend to be relatively stable; (2) when the number of training samples is constant, the same classification methods as well as the different classification methods have different degree of response to the training samples of the same quality grade; the degree of response of the same classification method to the different training samples quality level is also different.

HTML

Reference (15)

[1]	Ding Ling, Tang Ping, Li Hongyi. Dimensionality reduction and classification for hyperspectral remote sensing data using ISOMAP[J]. Infrared and Laser Engineering, 2013, 42(10):2702-2711. (in Chinese)
[2]	Duan Yunsheng, Zhang Dongyan, Huang Linsheng, et al. Comparison of hyperspectal and imagery characteristics of freezing stress and normal wheat[J]. Infrared and Laser Engineering, 2015, 44(7):2218-2223. (in Chinese)
[3]	Tao Qiuxiang, Zhang Lianpeng, Li Hongmei. The methods for selecting training samples in vegetation classification based on hyperspectral remote sensing[J]. Remote Sensing for Land Resources, 2005, 2(64):33-45. (in Chinese)
[4]	Bo Shukui, Ding Lin. The effect of the size of training sample on classification accuracy in object-oriented image analysis[J]. Journal of Image and Graphic, 2010, 15(7):1106-1111. (in Chinese)
[5]	Zhao Hui, Wang Yunjia. Research on the factors affecting the classification accuracy of ETM remote sensing image land cover/use[J]. Remote Sensing Technology and Application, 2012, 27(4):600-608. (in Chinese)
[6]	Zhu Xiufang, Pan Yaozhong, Zhang Jinshui, et al. The effects of training samples on the wheat planting area measure accuracy in TM Scale(I):the accuracy response of different classifiers to training samples[J]. Journal of Remote Sensing, 2007, 11(6):826-837.
[7]	Yan Jing, Wang Wen, Li Xiangge. Extracting the rice planting areas using an artificial neural network[J]. Journal of Remote Sensing, 2001, 5(3):227-231. (in Chinese)
[8]	Arai K. A supervised thematic mapper classification with a purification of training samples[J]. International Journal of Remote Sensing, 1992, 13(11):2039-2049.
[9]	Hui Wenhua. TM image classification based on support vector machine[J]. Journal of Earth Sciences and Environment, 2006, 28(2):93-95.
[10]	Wu Jianping, Yang Xingwei. Purification of training samples in supervised classification of remote sensing data[J]. Remote Sensing for Land Resources, 1996, 27(1):36-41. (in Chinese)
[11]	Nelson R F, Latty R S, Mott G. Classifying northern forests using Thematic Mapper simulator data[J]. Photogrammetric Engineering and Remote Sensing, 1984, 50(5):607-617.
[12]	Serra P, Pons X. Monitoring farmers' decisions on Mediterranean irrigated crops using satellite image time series[J]. International Journal of Remote Sensing, 2008, 29(8):2293-2316.
[13]	Zhang Shaojia. Remote sensing classification by combing multiple classifiers[D]. Changsha:Southeast University, 2010. (in Chinese)
[14]	Liu Li, Yu Qiang. A study on a classification method of remote sensing combined stratified classification with supervised classification[J]. Forest Inventory and Planning, 2007, 32(4):37-39. (in Chinese)
[15]	Zhang Hua. Study on reliable classification methods based on remotely sensed image[D]. Xuzhou:China University of Mining and Technology, 2012. (in Chinese)

Study on the effect of training samples on the accuracy of crop remote sensing classification

doi: 10.3788/IRLA201746.S126003

Abstract

References

Proportional views

通讯作者: 陈斌, bchen63@163.com

Article Metrics

Related

Proportional views