Enhanced Approach to Detect Ship Using Wavelet Transform

Synthetic Aperture Radar (SAR) generates highresolution images irrespective of weather condition, so it has been used in marine research. Nowadays, ship detection in marine application plays an important role. Ship detection becomes a cumbersome task due to the presence of Sea Clutters, Speckle noise and Land areas which increases the probability of false target detection. To solve this problem to some extent we deduced an approach to detect ships from highresolution Synthetic Aperture Radar (SAR) Images. This approach used a Land masked Image of coastal region having VH Intensity Band as an input later on HAAR wavelet Transform and Edge detection Method were applied on a landmasked image, which was then followed by a morphological filter to detect the ships. Here, the final result is validated by locating the geo-coordinates of the ships in the resultant image and an original image by overlaying them. Error Percentage for each pixel position using Mean Absolute Error is also calculated. The final Resultant Image is also overlaid on Google Maps, for ensuring the geo-referencing of Ocean/Sea Boundary is correct.