Noise Suppression Of Sonar Image Based On BM3D Algorithm

If human society wants to continue to develop and progress,it must have sufficient resources as the basic guarantee for development.At present,most of the resources are exploited on land,and many of them are non-renewable,so it is possible to turn to the sea.If you want to explore the ocean,sonar technology is a very good choice.However,because the complex noise environment of the ocean can seriously interfere with the image quality,denoising sonar images is particularly important.Although the BM3D image denoising algorithm is widely used,there are still some drawbacks.First,the display effect on detailed information such as the edge texture of the image is not good;Second,image denoising is inefficient.In view of the shortcomings of the above algorithm,an improvement plan is proposed,and the specific work content and innovation results are as follows:Firstly,aiming at the problem that BM3D algorithm denoising will destroy the edge texture and other detailed information of the image,a BM3D image denoising algorithm based on Canny operator edge detection is proposed.Firstly,the noise variance of the sonar image is estimated,and the variance is used as the parameter input of the algorithm,and then the edge detection algorithm based on the Canny operator is introduced to detect the edge of the image,extract and save the complete edge information of the image,and then do BM3D denoising processing on the image,and finally replace the image edge information after BM3D noise reduction with the saved complete image edge information,and finally obtain the BM3D denoising image based on edge detection.The results show that the improved algorithm saves the edge detail information of the image while realizing image denoising,and improves the denoising effect of the image.Secondly,aiming at the problem of low denoising efficiency of BM3D algorithm and combining the insensitivity of DBSCAN clustering to noise,a BM3D denoising algorithm based on DBSCAN clustering to achieve image segmentation is proposed.First,the sonar image is divided into a target area and a non-target area,the target area is the subject of image denoising,and most of the rest are non-target areas.The original BM3D algorithm was used to denoise the target area,and the median filtering algorithm with high denoising efficiency and good effect on marine speckle noise was used for the non-target area,and finally the denoising results of the two were combined to obtain the final denoising image.Since the search range of BM3D block matching is limited to the target area,the search time of block matching is greatly reduced,and the denoising efficiency of the image is improved.The results show that the improved algorithm not only improves the image denoising efficiency,but also improves the denoising quality.Finally,combined with the improvement ideas of the above two algorithms,a BM3D denoising algorithm based on image segmentation and edge detection is finally proposed.Firstly,the noise variance of the sonar image is estimated,and then the image is segmented into target area and non-target area,and the edge detection algorithm is introduced on the basis of BM3D denoising of the target area,and the median filtering is adopted for the non-target area.In order to further verify the effectiveness of the improved algorithm denoising effect,this paper compares with other classical denoising algorithms.The results show that the improved algorithm not only enriches the edge detail information of the image and improves the denoising quality,but also greatly improves the image denoising efficiency.