Design And Implementation Of Seabed Classification Software Based On Acoustic Image

With the development of marine information technology,the detection and development of marine resources has been paid more and more attention.People gradually abandoned the traditional manual sampling methods that took time and effort,and began to use the acoustic characteristics of seabed sediments to select more efficient acoustic remote sensing technology to achieve automatic classification of seafloor sediments.A large number of seabed backscattering intensity data can be obtained by the acoustic detecting device.The underwater acoustic image generated on this basis can provide an effective and reliable data source for the classification of the seabed,so the data post-processing software with the function of automatically dividing the seabed type is developed.It has an important meaning.This paper focuses on the feature extraction and classification recognition methods based on underwater acoustic images,and develops a classification and visualization software based on pattern recognition process and software engineering technology.Firstly,the imaging principle of two kinds of acoustic detection equipments,such as multibeam sounding system and side-scanning sound,is described in combination with the concept of acoustic-based seabed sediment classification.Secondly,various feature extraction and classification recognition methods based on underwater acoustic images are studied.The feature extraction method based on gray level statistics and gray level co-occurrence matrix and the classification method based on BP neural network algorithm are studied.Then,in the Visual Studio 2013 environment,the underwater acoustic image based on the development tools such as OpenGL and MFC is completed.The design and implementation of the classification software includes the software structure of the multi-document view application,the data acquisition module,the feature extraction module,the classification recognition module,and the 3D visualization module.Finally,by testing the functions of each module,it is verified that the software can complete the seabed.Topographic and geomorphic data reading,feature extraction and principal component analysis based on gray scale statistics and gray level cooccurrence matrix method,seabed sediment classification based on Back Propagation neural network algorithm,joint display of 3D seabed topography,sedimentary effect display,etc.Features.After a variety of experimental data processing and result analysis,this set of seabed classification software based on seabed acoustic images can achieve the expected requirements of software design in software function and algorithm effect,and can automatically and reliably complete the seabed quality.Classify and display tasks in 3D.