中国激光, 2012, 39 (11): 1114001, 网络出版: 2012-10-25   

基于区域增长的遥感影像视觉显著目标快速检测

Fast Detection of Visual Saliency Regions in Remote Sensing Image based on Region Growing
张立保 1,2,*
作者单位
1 北京师范大学信息科学与技术学院, 北京 100875
2 北京师范大学遥感科学国家重点实验室, 北京 100875
摘要
针对传统视觉注意模型在遥感影像视觉显著区域检测中存在的计算复杂度高、检测精度低等缺点,提出了一种新的视觉显著区域快速检测算法。首先利用整数小波变换降低遥感影像的空间分辨率,从而降低视觉注意焦点检测的计算复杂度;然后在视觉特征融合中引入二维离散矩变换,生成边缘与纹理信息更为丰富的遥感影像显著图;最后在显著图分析中提出区域增长策略来获得视觉显著区域的精确轮廓。实验结果表明,新算法不仅有效降低了遥感影像视觉显著区域检测的计算复杂度,而且能够精确描述视觉显著区域的轮廓信息,同时避免了对整幅遥感影像的分割与特征提取,为今后的遥感影像目标检测提供了一定地参考价值。
Abstract
The traditional visual attention model for the detection of visual saliency regions in the remote sensing image can lead to high computational complexity and low precision of detection. A new fast detection algorithm of visual saliency regions is proposed. The new algorithm firstly decreases the spatial resolution by integer wavelet transform, which can reduce the computational complexity of detection of the visual focus of attention. Then, the new algorithm proposes the two-dimensional discrete moment transform for visual feature fusion, which can generate the saliency map of the remote sensing image which has more abundant information of edge and texture. Finally, the region growing strategy based on the visual focus of attention is proposed in the saliency map analysis to acquire the precise contours of the visual saliency regions. The experimental results show that the new algorithm can not only effectively reduce the computational complexity of the detection of visual saliency regions in the remote sensing image, but also be able to accurately describe the contour information of visual saliency regions. In addition, it can avoid image segmentation and feature extraction for the whole image. The new algorithm provides a certain reference for the target detection of the remote sensing image in the future.

张立保. 基于区域增长的遥感影像视觉显著目标快速检测[J]. 中国激光, 2012, 39(11): 1114001. Zhang Libao. Fast Detection of Visual Saliency Regions in Remote Sensing Image based on Region Growing[J]. Chinese Journal of Lasers, 2012, 39(11): 1114001.

本文已被 7 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!