麻祥才,顾萍,钱志伟,肖颖,王晓红.基于HSI颜色空间噪声检测的彩色图像去噪算法[J].包装工程,2021,42(3):270-275.
MA Xiang-cai,GU Ping,QIAN Zhi-wei,XIAN Ying,WANG Xiao-hong.Color Image Denoising Algorithm Based on HSI Color Space Noise Detection[J].Packaging Engineering,2021,42(3):270-275. |
| 基于HSI颜色空间噪声检测的彩色图像去噪算法 |
| Color Image Denoising Algorithm Based on HSI Color Space Noise Detection |
| 投稿时间：2019-06-08 |
| DOI：10.19554/j.cnki.1001-3563.2021.03.038 |
| 中文关键词: 噪声检测 彩色图像 去噪算法 HSI颜色空间 |
| 英文关键词:noise detection color image denoising algorithm HSI color space |
| 基金项目:上海教育委员会和上海发展基金会“晨光计划”(18CGB09);“柔版印刷绿色制版与标准化实验室”项目(LGPSFP-01) |
| 作者 | 单位 |
| 麻祥才 | 上海出版印刷高等专科学校，上海 200093 |
| 顾萍 | 上海出版印刷高等专科学校，上海 200093 |
| 钱志伟 | 上海出版印刷高等专科学校，上海 200093 |
| 肖颖 | 上海出版印刷高等专科学校，上海 200093 |
| 王晓红 | 上海理工大学，上海 200093 |
|
| Author | Institution |
| MA Xiang-cai | Shanghai Publishing and Printing College, Shanghai 200093, China |
| GU Ping | Shanghai Publishing and Printing College, Shanghai 200093, China |
| QIAN Zhi-wei | Shanghai Publishing and Printing College, Shanghai 200093, China |
| XIAN Ying | Shanghai Publishing and Printing College, Shanghai 200093, China |
| WANG Xiao-hong | University of Shanghai for Science and Technology, Shanghai 200093, China |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 为了有效去除彩色图像中的椒盐噪声，提高彩色图像质量。方法 采用椒盐噪声检测和中值滤波相结合的方法，提出一种基于HSI颜色空间噪声检测的彩色图像去噪算法。将图像转换到HSI颜色空间，根据椒盐噪声在S通道具有极大值或极小值的特点判断出可疑椒盐噪声的位置，在H通道、I通道将可疑椒盐噪声分为噪声点和有用信号，对检测出的噪声像素进行中值滤波去噪。结果 采用文中算法去噪后，验证图像主观评价值(Z)为1.30，平均PSNR为37.54，SSIM为0.99，Entropy为7.31，在主客观评价上优于现在常用算法。结论 文中提出算法可以为彩色图像椒盐噪声的去噪提供理论基础，具有一定的实际应用价值。 |
| 英文摘要: |
| The work aims to effectively remove salt and pepper noise in color image and improve the quality of color image. A color image denoising algorithm was proposed based on HSI color space noise detection by combining salt and pepper noise detection with median filtering. The image was transformed into HSI color space, and the position of the suspicious salt and pepper noise was judged according to the characteristics of noise having maximum or minimum value in S channel. In H channel and I channel, the suspicious noise was divided into noise points and useful signals, and the detected noise pixels were denoised by median filtering. The subjective Z-score of denoising verification image was 1.30, the average PSNR was 37.54, SSIM was 0.99 and Entropy was 0.73, which was better than the commonly used algorithm in subjective and objective evaluation. The algorithm proposed can provide theoretical basis for the removal of salt and pepper noise in color images, and has certain practical value. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
渝公网安备 50010702501716号