An Efficient Algorithm for Boundary Detection

We proposed an efficient method to detect object and the boundary of the object. The object detection is done after the subtraction between background and object with background in gray scale wit threshold, which is generated from the normal distribution curve fitted histogram. In most of the existing methods for binary scale conversion and detection of object only one threshold used, here we used two threshold to get the better result while subtraction the object from background. We use our algorithm to get the best object image in binary scale and then find its boundary to get the shape of the object.