| 陈强,田杰,刘维,黄海宁,张春华.基于纹理特征的合成孔径声纳图像目标检测研究[J].声学技术,2013,32(4):273~276 |
| 基于纹理特征的合成孔径声纳图像目标检测研究 |
| Texture feature based target detection for SAS image |
| 投稿时间：2012-05-06 修订日期：2012-08-19 |
| DOI：10.3969/j.issn1000-3630.2013.04.003 |
| 中文关键词: 合成孔径声纳 目标检测 纹理特征 灰度共生矩阵 |
| 英文关键词: SAS(Synthetic Aperture Sonar) Target detection Texture feature Grey level co-occurrence matrix(GLCM) |
| 基金项目: |
|
| 摘要点击次数: 1442 |
| 全文下载次数: 1733 |
| 中文摘要: |
| 随着成像声纳技术的发展,声纳图像的目标检测与识别逐渐成为数字图像处理领域的一个重要研究课题。合成孔径声纳图像含有丰富的纹理特征,而灰度共生矩阵具有丰富的特征参数,可以从不同的角度对纹理进行细致刻画。采用灰度共生矩阵可以描述合成孔径声纳图像纹理方面的特征,通过计算灰度共生矩阵在方位向和距离向的能量、相关性、对比度和熵值,并构造特征向量,可以对合成孔径声纳图像中的目标进行准确检测。从实验结果可以看出,基于纹理信息可以准确实现合成孔径声纳图像的目标检测。 |
| 英文摘要: |
| With the development of image sonar technology, target detection and recognition for sonar image has become a more and more important research subject in the field of digital image processing. Synthetic aperture sonar (SAS) images contain rich texture features, and the gray level co-occurrence matrix (GLCM) that has rich characteristic parameters can characterizes the textures from different perspectives. So, GLCM is used to describe the texture features of SAS images in this paper, and the target in SAS images can be accurately detected by calculating the energy, correlation, contrast and entropy of GLCM and forming the characteristic vectors. From the results of experiments, it can be seen that the accurate target detection in synthetic aperture sonar image can be achieved based on texture information. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
