ABSTRACT
Image-object extraction is one of the most important parts in the image processing Object-based image analysis (OBIA) method classify individual pixels directly, This method first aggregates image pixels into spectrally homogenous image objects by using an image segmentation algorithm and then classifies the individual objects using any classifier. Image segmentation is one of the most important pre-processing stages in computer vision to classify similar pixels to one cluster. Clustered pixels have the same attributes such as texture, color and intensity. The most famous segmentation techniques are edge based techniques, region based techniques, clustering based techniques and watershed based techniques. To overcome non continuity and non differentiability there is an introduction of a new method based on the curvelet transform, which represents edge better than wavelet. My work is related to extracting objects from images by detecting edges based on the watershed and curvelet transform.
