QQ客服 官方微博 反馈留言

	Exploring Cross-Channel Texture Correlation for Color Texture Classification
	Qi Xianbiao; Qiao Yu; Li Chun-Guang; Guo Jun
	2013
会议名称	24th British Machine Vision Conference
会议地点	Bristol, ENGLAND
英文摘要	This paper proposes a novel approach to encode cross-channel texture correlation for color texture classification task. Firstly, we quantitatively study thecorrelation between different color channels using Local Binary Pattern (LBP) as the texture descriptor and using Shannon's information theory to measure thecorrelation. We find that (R, G) channel pair exhibits stronger correlation than (R, B) and (G, B) channel pairs. Secondly, we propose a novel descriptor to encode the cross-channel texture correlation. The proposed descriptor can capture well the relative variance of texture patterns between different channels. Meanwhile, our descriptor is computationally efficient and robust to image rotation. We conduct extensive experiments on four challenging color texturedatabases to validate the effectiveness of the proposed approach. The experimental results show that the proposed approach significantly outperforms its mostly relevant counterpart (Multi-channel color LBP), and achieves the state-of-the-art performance.
收录类别	ISTP
语种	英语
内容类型	会议论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/4485]
专题	深圳先进技术研究院_集成所
作者单位	2013
推荐引用方式 GB/T 7714	Qi Xianbiao,Qiao Yu,Li Chun-Guang,et al. Exploring Cross-Channel Texture Correlation for Color Texture Classification[C]. 见:24th British Machine Vision Conference. Bristol, ENGLAND.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。