传统三维矩不变量的计算方法基于区域特征,计算量较大。将基于区域的三维矩不变量计算转化到三维物体曲面轮廓上,构造了一种的新的曲面轮廓矩不变量。此不变量减小了计算复杂度,不仅具有平移、尺度和旋转不变性,而且独立于三维物体本身的颜色和灰度级。仿真实验表明,基于曲面轮廓矩不变量的识别算法对于三维目标具有良好的识别率,减小了计算复杂度。另外,对这种矩不变量在基于颜色、灰度等识别场合的应用做了推广。
Traditional approaches for 3D moment invariants computing are based on field features and cause high calculation cost.A series of novel moments named curved surface moments was constructed.Some properties of these new moments including the invariance on translation,rotation and scaling transforms were proved.Experiments indicate an encouraging high recognition rates without reducing the recognition performance.Some extensions of the moment invariants to 3D objects recognition based on the color and gray levels were introduced.