Sliding window based neural network for bottle mouth image segmentation

CORC > 厦门大学 > 信息技术－已发表论文

	Sliding window based neural network for bottle mouth image segmentation
	Fang, Li ; Xuan, Li ; Fan, Yang ; Yang F(杨帆)
刊名	http://dx.doi.org/10.4156/jcit.vol7.issue9.32
	2012
关键词	Image segmentation Neural networks Pixels Speech recognition
英文摘要	Image segmentation for bottle mouth is an important part of bottle mouth inspection system. In this paper, a sliding window based classification method was proposed for image segmentation in complex industrial environment. Firstly, the image segmentation problem is viewed as a classification task, and Backpropagation Neural Network (BP) is applied to classify each pixel from the target area according to the information extracted from the image. Different from the traditional image segmentation methods, an appropriate size of sliding window is used to extract pixel information. When the window slides, the image in the window is converted to a characteristics matrix which is used as the input of the BP neural network. Experimental results on real-world dataset showed that compared with traditional methods, the new method was more accurate and effective, especially in the conditions of complex background and external disturbance.
语种	英语
出版者	Advanced Institute of Convergence Information Technology
内容类型	期刊论文
源URL	[http://dspace.xmu.edu.cn/handle/2288/92798]
专题	信息技术－已发表论文
推荐引用方式 GB/T 7714	Fang, Li,Xuan, Li,Fan, Yang,et al. Sliding window based neural network for bottle mouth image segmentation[J]. http://dx.doi.org/10.4156/jcit.vol7.issue9.32,2012.
APA	Fang, Li,Xuan, Li,Fan, Yang,&杨帆.(2012).Sliding window based neural network for bottle mouth image segmentation.http://dx.doi.org/10.4156/jcit.vol7.issue9.32.
MLA	Fang, Li,et al."Sliding window based neural network for bottle mouth image segmentation".http://dx.doi.org/10.4156/jcit.vol7.issue9.32 (2012).