摘要
在小波变换的基础上,提出了一种基于小波变换的Wiener滤波去噪方法。采用该方法对苹果图像的加性噪声(高斯白噪声)去噪,结果PSNR为184.94,视觉效果清晰(而含噪声图像的PSNR为158.23,噪声太多不清晰),好于邻域平均法(PSNR为174.15,视觉上含部分噪声)、中值滤波法(PSNR为182.42)、小波阈值去噪(PSNR为171.59,视觉上也含部分噪声)和Wiener滤波去噪法(PSNR为173.65)的去噪结果,更好于数学形态学的去噪结果(PSNR为150.46,视觉上含较多噪声)。试验结果表明,基于小波变换的Wiener滤波方法对苹果图像加性噪声的去噪效果具有信噪比高、视觉效果好等优点。
A wavelet transform-based Wiener filtering method was put forward. The method was applied in reducing additive noise in apple images, as a result, PSNR was 184.94（visual effect was clear）, better than other methods such as neighborhood average （PSNR was 174. 15）, median filter （PSNR was 182.42）, wavelet thresholding de-noise （PSNR was 171.59）and Wiener filter （PSNR was 173.65）, and much better than mathematical morphology （PSNR was 150.46, with much noise in visual effect）. The experimental results showed that wavelet transform-based Wiener filtering method applied in reducing additive noise in apple image has the advantages of high signal-to-noise, better visual effect than conventional de-noise methods, wavelet thresholding de-noise and Wiener filter de-noise. So wavelet transform-based Wiener filtering method applied to reduce additive noise in apple image is effective and practicable.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2006年第12期130-133,143,共5页
Transactions of the Chinese Society for Agricultural Machinery
基金
西北农林科技大学优秀人才基金资助项目(项目编号:04ZR003)
校长基金资助项目(项目编号:08080101)
"985工程"首批青年学术骨干基金资助项目(项目编号:985-0568)
高校博士点基金项目(项目编号:20040712018)
