Research On The Underwater Integrated Navigation Algorithm Based On Side Scan Sonar Image

The rapid development of underwater autonomous underwater vehicle(AUV)has made the need for underwater integrated navigation technology more and more urgent.Traditional underwater combined navigation is the main technical means by inertial navigation and doppler(DVL)velocity measurement.Inertial navigation has the characteristics of high update rate and dynamic performance.However,pure inertial navigation has the characteristics of error over time,and the problem of low cost inertial measurement unit is more serious.The DVL meter is used as the auxiliary navigation system to modify the error of the inertial navigation system over time.However,DVL can be used in practical work because of the complicated underwater environment and AUV maneuvering(floating,diving,steering),etc.Some of the larger DVL coarse differences can be detected and removed directly according to AUV kinematics constraints or inertial navigation information.But when using low cost inertial navigation system(such as heading attitude reference system),inertial navigation provide information itself at the rate of accuracy is not high,the error growth is fast,conform to the AUV kinematics constraints,and close to normal speed value,duration long DVL gross error is often difficult to detect and eliminate.Side scanning sonar is one of the important equipment to detect the ocean,and the generated acoustic image can also provide auxiliary modification for underwater navigation.Based on the imaging features of the side scan sonar image,the lateral velocity of the AUV carrier coordinate system was extracted and applied to the detection and elimination of the DVL coarse data.In this paper,the application of side-scan sonar image in underwater navigation is studied.The following are some of the most innovative work in this paper:1.The side scan sonar image file is drawn to complete the image of the side scan sonar,and the noise reduction is processed.In this paper,a large number of DVL measured data are analyzed,and the correlation characteristics of the forward velocity and lateral velocity of DVL are obtained.2.Based on SIFT feature matching algorithm,the real-time feature spot discrimination of side-scan sonar images was performed.Feature points are mainly line characteristics,which can include water bottom line,surface line and water bottom feature contour line separated from the image.3.Based on SIFT feature matching algorithm,a real-time vector lateral velocity estimation algorithm is proposed.The algorithm estimated the lateral velocity of the carrier and the lateral velocity of DVL4.The autonomous navigation of underwater underwater vehicle navigation is used for the navigation of the combined navigation methods of the side scanning sonar,doppler velocimeter and inertial navigation device.The side scan sonar can generate the underwater acoustic image in real time.Aiming at the characteristics of underwater acoustic image data,the method of drawing underwater acoustic image is described.In order to incorporate the acoustic image feature matching in integrated navigation,this paper introduced SIFT feature matching algorithm according to the characteristics of the underwater acoustic image algorithm is improved,and introduces discriminant conditions based on least square method.After the matching discriminant conditions are introduced,the improved SIFT feature matching algorithm is improved greatly in the matching of the landmark,which avoids the occurrence of mismatching.Finally,the experimental data obtained by the side scan sonar equipment are verified by the algorithm,and the satisfactory results are obtained.5.The navigation algorithm based on INS / DVL combined navigation method codes the side-scan sonar image in geographic coordinates.