This report presents an algorithm for locating the cut points for and separating vertically attached traffic signs in Sweden. This algorithm provides several advanced digital image processing features: binary image which represents visual object and its complex rectangle background with number one and zero respectively, improved cross correlation which shows the similarity of 2D objects and filters traffic sign candidates, simplified shape decomposition which smoothes contour of visual object iteratively in order to reduce white noises, flipping point detection which locates black noises candidates, chasm filling algorithm which eliminates black noises, determines the final cut points and separates originally attached traffic signs into individual ones. At each step, the mediate results as well as the efficiency in practice would be presented to show the advantages and disadvantages of the developed algorithm. This report concentrates on contour-based recognition of Swedish traffic signs. The general shapes cover upward triangle, downward triangle, circle, rectangle and octagon. At last, a demonstration program would be presented to show how the algorithm works in real-time environment.