Abstract : In this paper, we investigate a technique, namely fuzzy Hough transform, to detect fuzzy geometrical entities like fuzzy lines and fuzzy circles. This technique identifies fuzzy lines and fuzzy circles by a voting procedure. This voting procedure is carried out in parameter space in the fuzzy set-theoretic framework. Sequentially, a concept of similarity measure between two fuzzy shapes is delineated. Moreover, we apply the proposed technique in a few authentic images to detect fuzzy lines and fuzzy circles. Also, the idea of similarity measure between two fuzzy shapes is implemented in two sample images of fuzzy shapes.