Improvement of contour extraction precision of active contour model with structuring elements

To achieve enhanced image recognition, it is necessary to accurately extract the contour of the recognition target from the input image in a preprocessing step. Contour extraction methods based on the active contour model(Snakes) require the operator to specify the parameter values that best catch the shape of the target. However, it is difficult to guess the parameter values since the relationship between target shape and values is unclear. One solution is the active contour model with structuring elements since its parameters are easily determinable. However, this method is not really practical since it fails if the boundary between the target and its background is not clear. Moreover, it is not efficient for natural images, since the backgrounds are not monotone. This paper introduces an improved contour extraction method based on the Sobel filter and thresholding.