Research On Preprocessing And Marine Target Detection Of Infrared Image Sequence

With the development of marine resources,marine economy develops rapidly,maritime rights and interests disputes are also gradually rising,and maritime traffic is becoming busier.Therefore,it is urgent to carry out real-time monitoring and control of its territorial waters.Infrared thermal imaging technology has the advantages of long work distance,high concealment,strong penetrating ability,strong anti-jamming ability,strong ability of target recognition,and all-weather operation.It has become one of the most important means to detect and identify the targets on the sea.In this paper,we mainly study the method of removing the periodic stripe noise in the infrared image,the method of extracting the sea area from the infrared image sequences,and Sea surface target detection method based on infrared image sequences.Specifically,major contributions of this dissertation are summarized as follows:1.Aiming at the problem of removing periodic stripe noise in infrared image,the spatial and frequency domain characteristics of the stripe noise are analyzed,and the comb-like characteristics of the stripe noise in the amplitude spectrum are derived.On this basis,an optimal method for estimating the size of sub-image is proposed.We first crop the original image into two sub-images with optimal size to make the comb-like impulse spectrum characteristics of stripe noise most prominent.In this condition,the noise patterns in sub-images are easier to remove.In addition,this paper also gives the design methods of two kinds of filters,an adaptive comb-type filter and local linear interpolation filter.The results indicate that the proposed method has a good performance in adaptability and stability for removing complex stripe noise in infrared images than others.2.Aiming at the problem of extracting sea area from the infrared image sequence,the characteristics of pure sea water background in the spatial and frequency domain are analyzed respectively in the horizontal and vertical directions.We find that the gray variance of sea water background in each direction and the mean value of frequency components in each direction have stable characteristics.On this basis,a new method of infrared image sea water area extraction based on the spatial frequency feature Gauss model is proposed.In this algorithm,the features of each direction are extracted first,and the principal component of the eigenvector is obtained by PCA algorithm.And then,the probability density estimation is performed by using the Parzen window function,the probability density function is fitted by a single Gauss function.Finally,the seawater background model is obtained and the background area of pure seawater is extracted.3.Aiming at the problem of sea target detection in infrared image sequences,a new algorithm of infrared target detection based on sea water background in two way half Fourier domain is proposed.In this algorithm,firstly,the background suppression filters are constructed according to the horizontal and vertical sea surface background model.Then,the filtered results are fused and enhanced.Finally,Otsu is used to extract the target.The experimental results show that the target detection algorithm is proposed in this paper has the following advantages: the area of error detection is less,the target area is more complete,and it is robust to non-uniform illumination.4.On the basis of the research of above,we design and implement two sea infrared sequential image processing software: the preprocessing software of infrared image sequence and the target detection algorithms verification software.