化石自动鉴定技术研究现状与展望  

Research Status and Prospect of Automated Fossil Identification Technology

作　　者：夏菁[1] 白志强[1] 王宝鹏[1] 吴一超 常洁琼[1] 

机构地区：[1]北京大学地球与空间科学学院,北京

出　　处：《地球科学前沿（汉斯）》2013年第3期159-164,共6页Advances in Geosciences

基　　金：国家自然科学基金项目(41172031);中国地质调查局地质调查项目(1212011085148,1212011120116)资助。

摘　　要：随着古生物学研究的深入与快速发展,传统化石鉴定方法效率低下、费时费力,已不能满足人们的需要。目前采用的化石自动鉴定技术主要有专家系统、多元统计分析和人工神经网络。近年来,随着计算机技术的发展,多元统计分析和人工神经网络已渐渐成为主流。二者的结合将是未来化石自动鉴定系统研发的必然趋势。The traditional identification method of fossils has always been tedious, inefficient, time-consuming and laborious. It cannot satisfy the needs of fast and further development of paleontological studies already. At present, there are three kinds of automation technologies used to identify fossils including expert system, multivariate statistical analysis and artificial neural network. With recent advancements in digital cameras and computer vision, multivariate statistical analysis and artificial neural network have become more and more popular and important. The combination of them will be an inevitable trend in the field of automated fossil identification systems in the future.

关 键 词：化石 自动鉴定 专家系统 多元统计分析 人工神经网络 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]