摘要
[目的]农业图像中难免存在噪声,噪声会干扰有用目标的识别。为了准确提取农业图像中的有用信息,拓展图像处理技术在农业工程中的应用,有必要去除农业图像中的噪声。[方法]本文将一种改进的中值滤波与视觉滤波LOG算子结合起来,用于对农业图像进行去噪。首先,通过LOG算子将图像分为边缘图像和非边缘图像,然后采用一种改进的加权中值滤波处理非边缘图像,边缘图像与处理后的非边缘图像融合为最终的去噪图像。[结果]利用本算法对农业图像进行测试,与传统的中值滤波算法进行对比,并采用峰值信噪比作为滤波算法性能的客观评价指标,本算法的峰值信噪比高于传统中值滤波算法8.15%,有更好的降噪效果。[结论]因此,该算法的滤波效果优于传统中值滤波算法,可有效去除农业图像中诸多因素产生的噪声。
[Objective]Agricultural images exist noise unavoidablyand noise can interfere with the identification of useful targets.In order to extract the useful information in agricultural images accurately and expand the application of image processing technology in agricultural engineering,it is necessary to remove the noise in agricultural images.[Methods]In this paper,the algorithm studied was to combine an improved median filter with the visual filter operator LOG for denoising agricultural images.Firstly,the image was divided into edge image and non-edge image by LOG operator.Then,an improved weighted median filter was used to deal with the non-edge image.Finally,the edge image was combined with the processed non-edge image as the final denoising image.[Result]Apply this algorithm to the agricultural images to test the performance of the filtering algorithm,compared with the traditional median filter algorithm,and use the peak signal to noise ratio as the objectivity evaluation,the proposed algorithm＇s peak signal-to-noise ratio was 8.15%,Was higher than the traditional median filtering algorithm,and had better noise reduction effect.[Conclusion]The algorithm of thismethod was better than the traditional median filtering algorithm,and could remove the noise generated by many factors in agricultural image effectively.
作者
李旭茹
李富忠
韩冬
徐晓宇
Li Xuru Li Fuzhong Han Dong Xu Xiaoyu(College of Software, Shanxi Agricultural University, Taigu 030800 , China College of Physical Information and Engineeingr , Shanxi Normal University, Lin fen 041000 ,China)
出处
《山西农业大学学报（自然科学版）》
CAS
2017年第9期670-673,678,共5页
Journal of Shanxi Agricultural University(Natural Science Edition)
基金
国家自然科学基金(61503231)
山西农业大学科技创新基金(2016009)
关键词
农业图像
LOG
滤波
去噪
Agricultural Images
LOG
Filter
Denoising
