首页 | 本学科首页   官方微博 | 高级检索  
     

红外图像增强方法的研究
引用本文:李云红,张龙,廉继红.红外图像增强方法的研究[J].西北纺织工学院学报,2010(4):516-520.
作者姓名:李云红  张龙  廉继红
作者单位:西安工程大学电子信息学院,陕西西安710048
基金项目:中国纺织工业协会科技指导性项目(2008012); 陕西省自然科学基础研究基金项目(2008JQ8022); 陕西省教育厅自然科学专项基金资助项目(09JK46810JK515); 西安工程大学2009年度校基础研究基金项目(09XG06)
摘    要:为解决对比度差、边缘细节模糊等问题,提出了用分段增强法提高红外图像对比度.分段小波增强算法能有效地同时增强红外图像对比度及边缘细节,并抑制噪声.建立了表征图像边缘的特征向量.根据红外图像边缘的特征,准确提取红外图像的边缘.通过神经网络边缘检测法对样本集训练,使网络具有依据边缘特征向量求解场景中物理边缘的能力.实验结果表明神经网络边缘检测算法的精度优于其他算法,抗噪声能力强、边缘定位能力强、检测精度高.

关 键 词:红外热像仪  图像增强  非均匀性校正  对比度增强  边缘检测

Research on the infrared image enhancement method
LI Yun-hong,ZHANG Long,LIAN Ji-hong.Research on the infrared image enhancement method[J].Journal of Northwest Institute of Textile Science and Technology,2010(4):516-520.
Authors:LI Yun-hong  ZHANG Long  LIAN Ji-hong
Affiliation:(School of Electronics & Information,Xi′an Polytechnic University,Xi′an 710048,China)
Abstract:For solving the problems of poor infrared image contrast and blurred edge details,segment enhancement method is proposed to enhance the contrast of infrared image.The wavelet is effective in enhancing the infrared image contrast and edge details and in reducing the noise.The characteristics of image edges may be displayed by characteristic vectors.According to characteristics of infrared image edges,the infrared image edges are correctly extracted.By training sample set,neural network has the ability to solve the scenery physical edges according to the edge characteristic vectors.Experimental result shows that the neutral network edge detection method is superior to other methods in noise resistance,edge positioning and detection accuracy.
Keywords:infrared thermal imager  image enhancement  nonuniformity correction  contrast enhancement  edge detection
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号