基于浓度特征向量的目标识别与分类

提出了一种新的基于零件浓度特征向量的目标识别与分类技术，以及由零件图像特征信息所构成的浓度特征信息的合理性验证方法。该技术不仅能够准确地反映目标图像的局部结构特征与整体结构特征之间的关系，而且较好地解决了计算机描述零件图像的特征信息的负担过重问题。实验表明，该技术具有识别准确、计算机负担小的优点。

Recognition and Classification of Objective Based on Density Feature Vectors

A new technique of object recognition and classification based on the density feature vector of the mechanical parts images is put forward, and a method for examinting the reasonability of the density feature information formed by the mechanical parts images is presented. The technique not only shows clearly the relation of the objective images between the partial structure features and the overall structure features, but an well solve the problem of the computer overload for describing the feature information of mechanical parts images. The experimental results show that the technique is excellence in recognizing to a nicety and lowering the computer load.