INTERNAL WAVE EXTRACTION IN SAR OCEANIC SURFACE IMAGE BASED ON STATIONARY WAVELET TRANSFORM
Mingxia He, Haihua Chen, Lingfei Guo, Xiaodong Zhuang, Hongping Li · 2006
An Internal Wave Feature Extraction （IWFE） algorithm based on Stationary Wavelet Transform (SWT) in SAR oceanic surface images is proposed. In this algorithm, SWT is applied to process SAR images and its coefficients are used to calculate the Wavelet Gradient Information (WGI). And the IWFE is obtained as edges by searching the calcaulated WGI’s maximum module value.. The morphological thinning algorithm and threshold technique are also used for edge refinement to suppress the other oceanic features, isolate Speckle noise. The results obtained from ERS-1 and ERS-2 SAR images show that the proposed method is more efficient than the traditional down sampling discrete wavelet transform in IWFE. The algorithm can be easily implemented and give an accurate and complete location of internal waves in SAR images. Oceanic internal wave is a general oceanic phenomenon. It is a very important research topic of oceanic dynamics and is close related to the ocean bionomics, ocean sedimentology and ocean physics. The observation of internal waves is very essential in ocean engineering, ocean navigation and undersea sound communication, etc. Internal wave travels within the interior of the sea, its space–time distribution is widely and diversified, hence, in-situ observation of internal waves is time consuming and expensive.