基于二维重组和动态窗格的水质检测紫外-可见光谱去噪算法

发布时间：2018-01-04 23:12

本文关键词：基于二维重组和动态窗格的水质检测紫外-可见光谱去噪算法　出处：《光谱学与光谱分析》2016年04期 　论文类型：期刊论文

【摘要】：立足于成功研制的紫外-可见光谱水质检测多参数测量系统,针对紫外-可见光谱水质多参数原位实时检测在精度、灵敏度、稳定性等方面的实际需要,开展了基于二维重组和动态窗格的水质检测紫外-可见光谱去噪算法的研究,以此提高紫外-可见光谱水质检测的测量精度。光谱法水质检测系统通常使用工业级低成本光谱仪,其输出光谱包含明显的非平稳噪声。传统去噪法难以在滤除噪声的同时保留谱线细节。而且,原位实时水质检测条件下,被测水样可能快速变化,传统去噪法中常用的多次采样求均值法将产生额外的测量误差。引入的去噪算法通过对水样光谱进行等间隙连续采样,将光谱数据张成由光谱轴和时间轴构成的二维矩阵,经过二维小波变换后,设置一个可变宽度的窗格在系数矩阵中水平滑动,使用窗格内的小波系数计算得到动态去噪阈值,并随窗格滑动构建去噪阈值向量,由此进行光谱去噪。其中,窗格宽度由相邻区域的噪声方差变化率决定,变化率较高的区域缩小窗格宽度,反之则扩大宽度。实验结果表明,这种去噪算法不仅能有效去除光谱中的非平稳噪声,而且能保留光谱的细节信息,有助于提高仪器的测量精度。与此同时,由于该算法并未使用时域平均,样本的快速变化对去噪性能的影响较小,适合在线或原位水质检测的水样本环境。
[Abstract]:Based on the successfully developed UV-Vis spectrum water quality detection multi-parameter measurement system, aiming at the actual needs of UV-Vis spectrum water quality multi-parameter real-time detection in precision, sensitivity, stability and so on. The research of UV-Vis spectral denoising algorithm for water quality detection based on two-dimensional recombination and dynamic pane is carried out. In order to improve the measurement accuracy of UV-Vis spectral water quality detection, the industrial low cost spectrometer is usually used in the spectral water quality detection system. The output spectrum contains obvious non-stationary noise. The traditional denoising method is difficult to preserve the spectral details while filtering noise. Moreover, under the condition of in-situ real-time water quality detection, the measured water sample may change rapidly. In the traditional de-noising method, the method of calculating the mean value of multi-sampling will produce additional measurement error. The proposed de-noising algorithm carries on the equal-gap continuous sampling to the water sample spectrum. The two-dimensional matrix composed of spectral axis and time axis of Zhang Cheng spectral data is set and a variable width pane is set to slide horizontally in the coefficient matrix after 2-D wavelet transform. The dynamic denoising threshold is calculated by using wavelet coefficients in the pane, and the threshold vector is constructed by sliding with the pane, and the spectral denoising is carried out. Where, the width of the pane is determined by the variation rate of the noise variance in the adjacent region. The experimental results show that the proposed denoising algorithm can not only effectively remove the non-stationary noise in the spectrum, but also retain the detailed information of the spectrum. At the same time, because the algorithm does not use the time-domain average, the fast change of the sample has little effect on the denoising performance, so it is suitable for the water sample environment of on-line or in-situ water quality detection.
【作者单位】：重庆大学光电技术及系统教育部重点实验室;重庆工业职业技术学院;
【基金】：国家自然科学基金项目(61401049) 四川省科技支撑计划项目(2012SZ0111) 重庆市研究生科研创新项目(CYS14039) 中国博士后科学基金面上项目(2014M560703) 重庆市博士后科研人员项目特别资助项目(Xm2014105)资助
【分类号】：X832;O657.3
【正文快照】： 引言水环境检(监)测可对水的质和量以及水体中影响生态与环境质量的各种人为和天然因素进行统一的定时或随时检测,它是环境保护的重要组成部分,关乎国家的可持续发展和人类生存环境的维护[1]。在线(on-line)或原位(in-situ)水质检测系统[2]可对污染源进行实时监控,及时准确的

【相似文献】