摘要
文章通过总结国内外现有的超新星搜寻方法,基于百度EasyDL定制化图像识别平台,提出了超新星搜寻系统设计。采用新疆星明天文台拍摄的可见光图像作为数据,对超新星自动搜寻系统的设计进行了研究,着重探讨了超新星特征提取、自动识别等问题。最终测试结果表明,此系统能迅速地缩小需要人工确认的可见光图像数量,从而准确地确定超新星。
This paper summarizes domestic and foreign methods of searching supernovas and presents the design of automatic searching system of supernovas based on EasyDL customized image recognition system.Visible light images are used photoed by Xinjiang Xingming Observatory as data,and the principle of design of automatic search of supernovas is studied,focusing on the problems of supernova feature extraction and automatic identification.The final results show that the system can rapidly reduce the number of images that require manual confirmation and precisely determine the supernovas.
作者
孙睿康
SUN Ruikang(Zibo Experimental High School,Zibo Shandong 255000,China)
出处
《智能计算机与应用》
2019年第1期262-263,共2页
Intelligent Computer and Applications
