This paper demonstrates the use of fuzzy sets for filtering contour lines in topographic map images. The fuzzy filtering is motivated by a need for effective automated analysis of data only available in the form of printed or scanned maps. The principal stages of the required pattern recognition and analysis procedures are briefly defined, and then the possibility of using fuzzy sets for filtering and segmentation is introduced. A filtering algorithm designed using fuzzy set theory is described. An implementation of the method using MATLAB, applied to a noisy topographic map image, is demonstrated. The described fuzzy filter achieves good noise removal which preserves and enhances the contour lines. The resulting image is suitable for input to the segmentation and classification stages.