Research Of Sonar Image Processing Based On Dual-core DSP

With the development of social demand and the rapid development of electronic products,digital signal processor(DSP)has become the first choice for engineering researchers in digital image processing,speech processing,radar,sonar and so on,with its high fixed point and floating point speed.The TMS320C6657,part of TI’s new C66 x family,is used in a wide range of engineering applications due to its powerful processing performance and low power consumption.In recent years,with the emphasis on marine development and protection of marine rights and interests in China,detection and identification of underwater targets is not limited to one-dimensional signals.High resolution imaging sonar has become an important means for people to explore marine knowledge,sonar image processing technology And target recognition technology has been the concern of scholars in various countries.In this paper,TMS320C6657 dual-core DSP to achieve sonar image processing algorithms,the main study of sonar image processing algorithms,dual-core DSP hardware platform development and sonar image processing algorithm of dual-core DSP to achieve three parts.Firstly,the sonar image processing method is studied,and the denoising effect analysis based on mean filter,median filtering and weighted median filtering is completed.In the sonar image thresholding method,the iterative method and the maximum interclass variance method are mainly studied,The morphological operation and the target extraction method are used to separate the different regions on the basis of segmentation.In order to better identify the image target,this paper focuses on the texture feature extraction method based on gray level co-occurrence matrix and the shape feature extraction method based on region invariant moment.In the second part,this paper introduces the TMS320C6657 chip features and the overall design of the system,describes the power supply module,clock module,peripheral interface module hardware circuit.Through the debugging of the system modules,the system performance indicators to achieve the desired requirements.At last,This paper studies the multi-core DSP software development method and the inter-core communication scheme.On this basis,the sonar image processing algorithm is implemented under the single-core serial framework.The program is carried out by allocating and using the memory and setting the compiler optimization level.optimization.Based on the parallel processing of the program,a multi-core parallel algorithm is designed.Compared with the serial algorithm,reduced task processing time,Algorithm performance has been significantly improved.