Research On Pseudo-color Image Processing In Side Scan Sonar

The side-scanning sonar detector is an acoustic detecting device widely used to detect underwater topography and is widely used in industrial and agricultural production and military fields.The side-scanning sonar device is a kind of gray-scale image generated by transmitting a pulsed sound wave signal to the bottom of the water,and then receiving the sound signal returned by the bottom of the seabed or the object,and the image is originally colorless,and the image is originally The recognition level of the grayscale image by the eye is only about 40,and the recognition type of the color can reach thousands of grades.Therefore,the pseudo-color processing can be performed on the sonar image in the later processing to improve the image of the side scan sonar device for the human eye.Recognition rate.This study is aimed at the research status and shortcomings of the side-scanning sonar technology at home and abroad.The original data of the side-scanning sonar has large noise and narrow gray-scale distribution.By comparing the more effective de-drying algorithm,BM3 D is selected as the denoising algorithm.Corresponding improvements have been made to improve the processing speed.For the problem that the gray scale of the sonar image is narrow,the histogram equalization is first used to pull the gray level of the image to a relatively wide range to prepare for the subsequent pseudo color coding.The paper then proposes a pseudo-color coding method based on human perception of color difference LCH space based on the calculation formula of human eye perception color difference.Based on this method,combined with the important characteristics of sonar image edge information,The image is subjected to multi-threshold segmentation,and the image is separately subjected to edge pseudo color in different segmentation layers and pseudo color is performed on the original grayscale image on the other layer,and then the pseudo image is merged.Considering that the sonar image can be divided into the classification characteristics of the target area,the background area and the shadow area,it is proposed to use the FCM algorithm combined with the image texture to cluster the images.After clustering,the different classes are converted by the histogram intersection method.To make the image richer and more natural.Through the method proposed in the paper,the peak signal-to-noise ratio of theimage after the BM3 D noise reduction can reach 34.33,and the MSE can reach0.1474.The image entropy of the image can be compared with the processing method based on the human eye recognition rate.The original grayscale image is increased by2 times.The CIDE2000 parameter can also be increased by 1.83 times relative to the original grayscale image.Image sharpness can be increased by 3.73 times relative to the original image.This paper proposes that the research method can significantly improve the human eye recognition rate of the image,and the sharpness of the image is also greatly improved.The method used in the paper has greatly improved the imaging level of the side scan sonar.