Research On Feature Extraction And Matching Method Of Side-scan Sonar Image

With the rapid development of contemporary science and technology,people are exploring more and more unknown areas,and have begun to increase research on space and oceans.The abundant natural resources in the ocean are more precious under the condition of the continuous consumption of land natural resources.On the other hand,in the military competition of all countries in the world,the marine strategy also occupies an important position.Therefore,all countries are increasing the exploration and development of the ocean.Because of the complex marine environment,conventional optical instruments can not be used normally in the marine environment.At present,the detection of seabed topography and targets mainly depends on sonar system,because sonar system uses sound waves to detect the seabed,which is less affected by the marine environment.In this paper,a side scan sonar system is used to obtain the seabed sonar map,and the seabed information is extracted by analyzing the sonar image.This paper mainly studies the analysis and processing of side scan sonar data,the feature extraction of sonar image and the matching of sonar image features.The main research process is as follows:Firstly,the working principle of the side scan sonar system is introduced,and the acquired scan data is analyzed to obtain the side scan sonar image to be processed.Since the scanning sound waves are affected by the complex environment of the ocean in the water,the final sonar map contains more noise.For this situation,a variety of filtering methods are used to preprocess the sonar map,and the image effects after filtering are compared and analyzed.Complete the denoising of the sonar image.Secondly,feature extraction is carried out for the pre-processed sonar image.In this paper,the feature points and line features are studied.The robust SIFT algorithm is used to extract the feature points of sonar image,and the description vectors are constructed to prepare for the matching of feature points.Hough transform and LSD algorithm are used to extract linear features from sonar images.At the same time,some problems in Hough transform space are improved.In order to improve the quality of the extracted straight line,an improved fitting method is designed in this paper,which can effectively fit the straight line features.Then,the matching of the features is completed on the basis of feature extraction.The constructed description vectors are used to match the feature points between images,and the feature points after initial matching are screened by RANSAC(Sampling Consistency).Constructing the matching condition vector,using some invariants in affine transformation tocomplete the matching between lines;introducing projective transformation,on the basis of mapping,constructing the matching condition of line feature from angle and position,and applying the condition to the matching of feature lines.The experimental results prove the feasibility of the proposed method,which shows that the proposed method has good linear feature matching performance.Finally,the simplified research on the feature of the extraction line is carried out.The extraction of the front line features reveals that there are too many similar line features,which will have certain influence on the description and matching results of the image information.Therefore,there are too many similar features for the line features.In this paper,the corresponding line feature simplification method is designed to deal with it.The experiment shows that the simplified line feature expresses the image information better and the line feature matching rate also improves.