The Edge Detection Of Sonar Image Based On Dual Tree Complex Wavelet Transform

The sonar image edge detection plays a very vital role in sonar image analysis. It aims to extract the bright and dark regions from the complex sea floor reverberation area of the target, and it can preserve the edge information of the original image to a certain extent. The identification of objects in the ocean, target tracking and classification mainly depends on whether we can accurately on the sonar image edge detection. It has extensive application that the sonar image edge detection in the field of military and civil fields.The dual tree complex wavelet transform not only adhering to the characteristics of wavelet transform for instance time-frequency localization analysis ability, perfect reconstruction, limited redundancy and efficient computing ability, but also to improved the defects of wavelet transform for approximate shift invariance and good directional selectivity. So it is widely used in image processing such as edge detection.In this paper I have use Lee filtering method of sonar image to filtered firstly and then dual tree complex wavelet decomposition to the filtered image. Next processing to coefficients and reconstruction of image to extract the edge. Finally thresholding and refine image to optimize the edge. In addition to the above methods, this paper also combine with the dual tree complex wavelet and Canny to edge detection. Soft thresholding process after dual tree complex wavelet decomposition of sonar image and then edge detection use Canny. The result show that the extracted edge so accurate positioning in these way, detection effect is also improved significantly.