论文部分内容阅读
从人类手印中提取信息确定性别以推断史前手印岩画中男女的性别角色。人们曾以手工的方式测算手指长度的比例和其他物理特征来判断性别角色。绝大多数传统的研究方法是基于手工测量长度，因此常常受制于出版的图片中信息量的缺失。我们已经探索了一种通过测量获取手部图像的信息并基于现代机器学习确定性别角色。这是目前已知的用自动化程序代替耗时的手工测量以此来确定史前岩画作者性别身份的方法。我们的研究为两性异形和旧石器时代晚期社会的劳动分工提供了定量的实证。此外，除了分析历史中的手印图像，该研究方法也具有应用于犯罪取证和人机交互方面的潜在可能。“,”Information was extracted from human fingerprints to determine gender to infer gender roles of men and women from prehistoric fingerprints. The manual way to measure the proportion of finger length and other physical characteristics were once used to determine the gender roles. The vast majority of traditional research methods were based on manual measure, therefore it is often subject to the deficiency of the amount of information published in the picture. Based on modern machine, we have explored a method of obtaining hand image information via measure to determine gender roles. It is now known automated procedures to replace time-consuming manual measurement in order to determine the authors’ gender identity of prehistoric rock art. Our research provides quantitative evidence to sexual dimorphism and the division of labor in the late upper-paleolithic societies. In addition to analysis of fingerprint image in the history, the study also has the potential to be applied into criminal forensics and human-computer interaction aspects.