摘要
对现行各种γ能谱滤波方法进行比对研究,发现各种滤波方法对能谱数据均有一定压缩与展宽;能域滤波时,谱线展宽增加重峰出现几率;频域滤波时,应注意滤波器及其参数的选取;小波变换能很好地保持信号的高频突变成分不受破坏,相比较而言,改进阈值方法能很好地保留硬阈值与软阈值的优点,适于弱峰的检测.针对重心法滤波器的不足,提出一种1倍标准差检测滤波算法,该方法能很好地保留能谱突变成分,维持谱峰净面积基本不变,对平台滤波效果等同于5点重心法,且算法简单.
Comparative study was conducted on results of gamma-ray spectrum by using a majority of active smoothing method, which were used to show filtering effect. The results showed that peak was widened and overlap peaks increased with energy domain filter in γ ray spectrum. Filter and its parameters should be seriously taken into consideration in fre- quency domain. Wavelet transformation can keep signal in high frequency region well. Im- proved threshold method showed the advantages of hard and soft threshold method at the same time by comparison, which was suitable for weak peaks detection. A new filter was put forward to eke out gravity model approach, whose denoise level was detected by stand- ard deviation. This method not only kept signal and net area of peak well, but also attained better result and had simple computer program.
出处
《南华大学学报（自然科学版）》
2011年第2期1-5,共5页
Journal of University of South China：Science and Technology
基金
江西省教育厅青年科学基金资助项目(GJJ10176)
核技术应用教育部工程研究中心(东华理工大学)开放基金资助项目(HJSJYB2010-07)
关键词
改进重心法
小波分析
傅立叶变换法
改进阈值
improved gravity model approach
wavelet analysis
fourier transform method
improved threshold
