可变形模型中形状参数对模型形状的影响

Configuration on Model Shapes by Adjusting Model Parameters

在图像处理和计算机视觉研究领域中,目标物体的定位始终是一个值得探讨的问题。论文以一组枫叶图像定位实例为背景,介绍了一种基于统计形状模型的主动形状模型(ASM)进行物体目标定位的理论框架。论文主要介绍了利用主分量分析方法构造先验模型,再通过调整先验模型中的参数,来观察模型的变化情况。论文以枫叶图像为训练样本,进行了大量的实验,实验表明,可变形模型中形状参数对模型形状的影响是有规律可寻的。

Localization of an object in an image is always one of the most typical research topics in the fields of image processing and computer vision.With the object localization of maple leaves as the application background,a statistical shape model based framework of Active Shape Model（ASM）is presented.In this paper,the effect is analyzed by adjusting the model parameters of objects.Those results show that the effect by adjusting the model parameters of objects is regular.