摘要
实验记录镀锌钢板和冷轧钢板在不同实验环境下的变色痕迹,分析6个主要因素对其的影响,如材料类型、加热方式、加热温度等。使用颜色特征和梯度特征对金属变色痕迹图像进行特征提取和表达。使用支持向量机方法进行金属变色痕迹图像分类识别。在此基础上编写程序对金属变色痕迹图像进行识别分类。实验结果表明,该方法能够实现对金属变色痕迹特征的快速准确识别。
The color trace of galvanized steel plain sheet and cold rolled steel plate under different test environment was tested and recorded. Six factors were analyzed, such as material type, heating method and heating temperature. The features of medal discoloration mark were extracted and expressed by color and gradient feature. Using support vector machine, the discolora- tion mark images were classified and recognized. Program was written to identify and classify the medal discoloration trace im age. Tests showed that, the method can identify the medal dis- coloration trace feature fast and accurately.
出处
《消防科学与技术》
CAS
北大核心
2016年第12期1769-1772,共4页
Fire Science and Technology
基金
公安部技术研究计划项目(2014JSYJA010)
2016中国消防协会科学技术年会"青年消防学者论坛"交流论文
关键词
金属变色痕迹
计算机视觉处理
支持向量机
火灾调查
medal discoloration trace
computer visual process-ing
support vector machine
fire investigation
