Under consideration are the scalable masks for detection of corner structures in digital images which are used when processing by a window sliding through an image. The proposed matrices of masks of arbitrary size are constructed by adding rows and columns along the perimeter to the matrix of smaller masks. The submatrices remain unchanged, whereas some new elements are added by repeating the submatrix entries that preserve the structure of the corner. The algorithm can be used in processing visual data of robotics, aerial photography, and crystallography.

中文翻译：

---

利用可扩展蒙版检测图像中的角点结构

在考虑中的是用于检测数字图像中的角结构的可缩放掩模，当通过滑动穿过图像的窗口进行处理时使用可缩放掩模。通过将沿周边的行和列添加到较小的蒙版矩阵中，可以构造任意大小的蒙版矩阵。子矩阵保持不变，而通过重复保留拐角结构的子矩阵条目来添加一些新元素。该算法可用于处理机器人技术，航空摄影和晶体学的视觉数据。