| 多视角多尺度陨石坑特征检测识别方法 |
| |
| 引用本文： | 高锡珍,汤亮,崔平远.多视角多尺度陨石坑特征检测识别方法[J].宇航学报,2022,43(7):921-929. |
| |
| 作者姓名： | 高锡珍 汤亮 崔平远 |
| |
| 作者单位： | 1. 北京控制工程研究所,北京 100094； 2. 空间智能控制技术重点实验室,北京 100094；3. 北京理工大学深空探测技术研究所,北京 100081 |
| |
| 基金项目： | 国家重点研发计划科技创新2030-“新一代人工智能”重大项目(2018AAA0102700)； |
| |
| 摘 要： | 针对序列图像间视角尺度变化大带来的特征无法检测和识别效率低等问题，提出了多视角多尺度陨石坑特征检测识别方法。基于明暗区域信息和形态学处理实现陨石坑图像特征边缘粗提取，利用快速傅里叶变换计算模板与图像特征的匹配度，从而实现图像特征所在区域粗定位。在此基础上，通过引入惯性测量信息计算图像间重叠面积，更新模板形状，并预测搜索区域，解决了序列图像间视角尺度变换及信息冗余带来的特征无法检测和识别效率低的困难，最终在单一陨石坑区域内实现其精细检测。
|
| 关 键 词： | 地外探测 陨石坑 特征检测 特征识别 多视角多尺度 |
| 收稿时间： | 2021-10-26 |
Detection and Recognition Method for Multiview and Multiscale Crater Features |
| |
| GAO Xizhen,TANG Liang,CUI Pingyuan.Detection and Recognition Method for Multiview and Multiscale Crater Features[J].Journal of Astronautics,2022,43(7):921-929. |
| |
| Authors: | GAO Xizhen TANG Liang CUI Pingyuan |
| |
| Affiliation: | 1. Beijing Institute of Control Engineering, Beijing 100094, China; 2. Key Laboratory of Space Intelligent Control Technology, Beijing 100094, China; 3. Institute of Deep Space Exploration, Beijing Institute of Technology, Beijing 100081, China |
| |
| Abstract: | Aiming at the problem that the features can not be detected and the recognition efficiency is low due to the large range of viewing angle scale changes between sequence images, a multi view and multi scale crater feature detection and recognition method is proposed. The feature edges of crater image can be extracted roughly based on the information of light and dark regions and morphological processing. The fast Fourier transform is used to calculate the matching degree between the template and the image feature, so as to realize the rough location of the image feature region. On this basis, by introducing inertial measurement information, the overlapping area between images is calculated, the template shape is updated, and the search area is predicted. Therefore, the difficulty of feature detection and low recognition efficiency caused by the view and scale distortions and information redundancy of sequence images is solved. At last, the fine detection is realized in a single crater region. |
| |
| Keywords: | Extraterrestrial exploration Crater Feature detection Feature recognition Multiview and multiscale |
|
| 点击此处可从《宇航学报》浏览原始摘要信息 |
|
点击此处可从《宇航学报》下载全文 |
|
