摘要
金标试纸定量分析设备采用的去噪算法对金标试纸进行降噪处理效果的优劣,直接影响到金标试纸定量分析的准确度。为了更好地解决金标试纸图像的噪声问题,结合各向同性和异性的扩散原理改进分裂Bregman迭代的全变分算法,利用该算法对林可霉素金标试纸噪声图像进行降噪处理。实验表明,与基于分裂Bregman各向同性全变分(SBITV)、基于分裂Bregman各向异性全变分(SBATV)以及金标试纸定量分析设备最常用的中值滤波算法相比,文中改进的算法克服了SBITV导致的图像边缘模糊问题以及SBATV产生的阶梯效应,既消除了噪声也保留了金标试纸图像的细节信息,获得了优越的降噪性能。在三种噪声的干扰下,本文改进的算法PSNR值比SBITV的PSNR值提高2dB以上;比中值滤波PSNR值至少提高15dB;比SBATV的PSNR值至少提高20dB。
The denoising algorithm used by the gold standard test paper quantitative analysis equipment directly affects the accuracy of the quantitative analysis of the gold standard test paper.In order to better solve the noise problem of gold standard test paper imag-es,this paper combines the isotropic and anisotropic diffusion principle to improve the total variational algorithm that splits Breg-man iteration,and uses this algorithm to reduce the noise image of lincomycin gold standard test paper.Experiments show that com-pared with the most commonly used median filtering algorithms for split-Bregman isotropic total variational(SBITV),split-Breg-man anisotropic total variational(SBATV)and the median filter algorithm used by gold standard test paper quantitative analysis equipment,the improved algorithm overcomes the problem of blurred image edges caused by SBITV and the step effect generated by SBATV,which not only eliminates noise but also retains the detailed information of gold standard test strip images,and obtains superior noise reduction performance.Under the interference of the three kinds of noise,the PSNR value of the improved algorithm is increased by more than 2dB compared with the PSNR value of SBITV,at least 15dB higher than the median filtered PSNR value,and at least 20dB higher than the PSNR value of SBATV.
作者
贺凤娟
陈伟利
HE Feng-juan;CHENWei-li(College of Electrical Engineering and Computer,Jilin Jianzhu University,Changchun 130118,Jilin)
出处
《电脑与电信》
2023年第7期80-85,共6页
Computer & Telecommunication
作者简介
贺凤娟(1997—),女,安徽阜阳人,硕士研究生,研究方向为嵌入式系统设计、图像处理。
