论文部分内容阅读
本研究借助计算机技术，初步探讨鼠类头骨自动分科的识别模型。研究选择4个科的8种鼠的成年组作为研究对象，用BugshapeV1.0提取数学形态特征。对这些特征进行比值化处理后分析建模，用于鼠类的分科识别。从交叉判别结果来看，利用数学形态比值化特征建立的模型可以进行很好地识别，但要提高识别率或再增加鼠种时，需要进一步提取新特征并优化。“,”This study focused on modelling explicit automatic recognition models with the aid of computer technology. The study selected the adults group of 8 kinds in 4 families as samples, mathematical morphological characteristics were extracted with BugshapeV1.0. Did analysis of the characteristics of ratio and modeling, branch identification for rats and mice. The cross validation results showed that models built by mathematic morphological characters can well identify rats. However, it needs more features and optimize them to improving the accuracy or increasing rat varieties, need to further extract new characteristics and optimization.