摘 要:图像在进行传递表达信息时会受到一定噪音的影响,使得图像的信息度降低,整个图像的平滑度降低,所以,为了保证图像在传输过程中的准确度、图像整体的效果,会对图像进行预处理——专门的去噪,专门的去噪过程有多种,其中包括空域、变换域和机器学习三大类方法,而其中较为成熟的应用技术要属机器学习方法,机器学习的图像去噪又有三大类,分别是神经网络、稀疏算法和向量算法,通过对去噪过程进行研究,提高图像在信息传递中的清晰度和准确度,同时,增强去噪的作用效果。本文将主要就机器学习的去噪过程进行研究,分别从机器学习的去噪研究的含义,去噪研究的应用及提高去噪作用效果的相应措施几个方面进行详细讨论,为日后相应的措施改变提供理论参考和借鉴。
关键词:机器学习;图像去噪;应用改进
中图分类号:TP391.41;TP181 文献标识码:A 文章编号:2096-4706(2019)14-0071-03
Research on Image Denoising Based on Machine Learning
CHEN Qi,ZHANG Yuehua,WANG Hong
(Shandong Huayu University of Technology,Dezhou 253034,China)
Abstract:Image in conveying information will be influenced by a certain noise,make the image information degree is reduced,the smoothness of the image is reduced,Therefore,in order to guarantee the accuracy of image in the process of transmission,the effect of the image as a whole,the image preprocessing,through specialized denoising,there are many types of specialized denoising process,including the airspace,transform domain and machine learning methods,and the more mature application technology is machine learning methods,machine learning image denoising and three categories,respectively is neural networks,sparse algorithm and vector algorithm,through the research of the denoising process improve the clarity and accuracy of the image in the transmission of information,and enhance the effect of denoising. This paper will mainly study the denoising process of machine learning,and discuss in detail the meaning of the denoising study of machine learning,the application of the denoising study and the corresponding measures to improve the effect of the denoising study,so as to provide theoretical reference and reference for the corresponding measures in the future.
Keywords:machine learning;image denoising;application of improved
基金项目:山东华宇工学院科技计划项目:基于深度学习的图像去噪研究(项目编号:2018KJ08);2016 山东省本科高校教学改革研究项目:基于CDIO 的校企合作应用型本科网络工程专业人才培养模式研究与实践(项目编号:C2016M078)。
参考文献:
[1] 江唯奕. 数字图像去噪算法原理及应用 [J]. 电子制作,2019(6):26-28.
[2] 李胜旺,韩倩. 基于深度学习的图像处理技术 [J]. 数字技术与应用,2018,36(9):65-66.
[3] 王萌萌,屈红伟,孙燕,等. 基于自适应非局部稀疏编码图像去噪方法 [J]. 计算机工程与设计,2017,38(8):2178-2183.
[4] 徐政超. 使用稀疏表达和机器学习的行人检测技术探讨 [J]. 科技创新与应用,2017(10):75.
[5] 郭德全,杨红雨,刘东权,等. 基于稀疏性的图像去噪综述 [J]. 计算机应用研究,2012,29(2):406-413.
作者简介:陈琦(1985-),男,汉族,山东济宁人,教师,初级职称,硕士,研究方向:计算机应用技术。