论文部分内容阅读
矩作为描述图象的一种方式已在图象处理和模式识别过程中得到应用。本文基于 Hu 所定义的函数矩的方式提出了倒矩概念。文中列出了倒矩具有的一些性质。文中推出倒矩和中心矩之间的关系。从倒矩和中心矩之间的关系以及倒矩的性质中可以看到函数的一个特性在低阶倒矩得到反映后,相应它就不影响高阶倒矩,即一个特性仅在一个倒矩中得到反映。Hu 矩却没有这个特性。一个特性在低阶矩中出现后。它仍然在高阶矩中起作用。规一化过程只能消除部份特性的影响。因此可说,例矩是描述图象的一个较好参量,文中同时对物体旋转后的倒矩以及倒矩应用作了一些讨论。
Moments as a way of describing images have been used in image processing and pattern recognition. In this paper, the concept of torque is proposed based on the method of moment defined by Hu. The article lists some of the properties that the torque has. The paper introduced the relationship between torque and center moment. It can be seen from the relationship between the moments of torque and the moments of the center and the nature of the moments that a characteristic of the function is reflected in the moments of lower order inversions and accordingly it does not affect the moments of higher order inversions, Be reflected. Hu Mo did not have this feature. After a feature appears in the lower moments. It still plays a role in higher moments. The normalization process can only eliminate some of the characteristics of the impact. Therefore, it can be said that the moment is a good parameter to describe the image. In the meantime, some discussions are made on the moment and the moment of rotation after the object is rotated.