Research on Image Classification Method Based on Adaboost-DBN

Sun, Huadong and Tao, Wuchao and Wang, Ran and Ren, Cong and Zhao, Zhijie (2019) Research on Image Classification Method Based on Adaboost-DBN. In: Wireless and Satellite Systems. 10th EAI International Conference, WiSATS 2019, Harbin, China, January 12–13, 2019, Proceedings, Part I.

[thumbnail of 31599.pdf] PDF
31599.pdf

Download (556kB)

Abstract

Image classification has been applied in many fields, which is an important branch of computer vision and pattern recognition. The boosting algorithm which is belong to ensemble learning can integrate several homogeneous classifiers, and combine the output layer’s result of every classifier to improve the final classification accuracy. In this paper, the Adaboost-DBN algorithm is used to combine the four weak classifiers (DBN) and construct a strong classifier. The Adaboost-DBN algorithm is based on the Adaboost M1 algorithm and is used to achieve higher classification accuracy. The proposed algorithm is tested on the Corel-1K data set, and the result of classification is significantly improved comparing to other classifiers.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Date Deposited: 04 Mar 2026 12:02
Last Modified: 20 Nov 2026 11:55
URI: https://eprints.eai.eu/id/eprint/22409

Actions (login required)

View Item
View Item