一种视频监控图像条纹噪声去除算法

针对视频图像中图像传感器故障引起的等间距横纹噪声去除问题。提出了基于陷波滤波器的去除条纹噪声算法;首先将视频条纹图进行傅里叶变换,通过频域累积分布函数映射法构造自适应滤波器,用该滤波器对变换后的条纹图像进行滤波,最后用Laplacian算子进行锐化处理得到最终去噪声图像。实验结果表明,本算法适合用于去除有图像传感器故障引起的等间距横纹噪声。

A stripe noise removal method for the video surveillance image

For the evenly spaced horizontal noise removal problem caused by image sensor fault in video images,we propose a stripes noise removal algorithm based on depression wave filter.First of all,we make Fourier transformation for video stripe images,and construct an adaptation filter based on depression filter through cumulative distribution function map method in frequency domain.Then we make the inverse Fourier transformation for the filtered video stripe images and get the restoration image.Finally,we sharpened the restoration image using Laplacian operator and overlay it to the restoration image,then we get the final noise removal image.Experimental results show that,the proposed algorithm is suitable for spaced horizontal noise removal caused by the image sensor fault.