| ID | 原文 | 译文 |
| 55137 | 首先,利用改进的比例边缘检测算子获得初始过分割结果;然后,设计一个基于相邻区域的面积和边缘信息的区域合并优先级函数来引导区域合并的进行,该方法提高了模型参数的估计精确,同时保留图像的强边缘;最后,将边缘信息区域合并技术用于求解基于多边形网格的最短描述长度准则SAR图像分割模型。 | First of all, the proportion of using improved edge detection operator to obtain the initial segmentation result;Then, design a based on the area of the adjacent area and edge information of the regional consolidation priority function to guide regional merge, the method improves the model parameters estimation precision, while preserving the image of strong edge;Finally, the edge information region merging technology is used for solving minimum description length criterion based on polygon mesh model of the SAR image segmentation. |
| 55138 | 实验表明,与同类方法相比,本文方法的边缘检测能力与定位精度均有提高。 | Experimental results show that compared with other methods, the method of edge detection and positioning accuracy are improved. |
| 55139 | 提出了一种基于目标高分辨距离像时频域非负稀疏编码的合成孔径雷达(synthetic aperture radar,SAR)目标识别方法。 | Is put forward based on the target high resolution range as time and frequency domain non-negative sparse coding of synthetic aperture radar (synthetic aperture radar, SAR) target identification method. |
| 55140 | 首先,将目标的SAR复图像转换为高分辨距离像。 | First of all, the target SAR complex image converted to high resolution distance. |
| 55141 | 然后,采用自适应高斯基表示方法计算每个距离像的非负时频矩阵。 | Then, using the adaptive gaussian said method to calculate each distance as non-negative time-frequency matrix. |
| 55142 | 其次,对训练目标所有距离像的时频矩阵采用非负稀疏编码方法学习时频字典。 | Second, the training goal and all the distance as the time-frequency matrix by non-negative sparse coding method to study the time-frequency dictionaries. |
| 55143 | 在目标识别中,通过将每个距离像的时频矩阵投影到低维的时频字典上来提取特征矢量。 | In target recognition, through each distance as the time-frequency matrix projection to low dimensional time-frequency dictionaries extracted characteristic vector. |
| 55144 | 最后,在提取特征矢量的基础上,通过支撑向量机目标识别决策实现目标识别。 | Finally, on the basis of the feature vectors extracted by support vector machine (SVM) target recognition decision target recognition. |
| 55145 | 采用美国"运动和静止目标获取与识别计划"公开发布的SAR图像数据库进行算法验证实验。 | Using the "motion and stationary target acquisition and recognition programs" public release experiment verifies the algorithm of SAR image database. |
| 55146 | 实验结果说明了提出方法的有效性。 | The experimental results show the effectiveness of proposed method. |