2005 年 59 巻 4 号 p. 610-613
In conventional fuzzy Hough transforms, lines are detected using a sequential iteration method that deletes data points on the detected lines. Detecting another line intersecting the detected lines is difficullt because some data points have been deleted in the previous process. We propose a fuzzy Hough transform method that preserves all the data in the iterative detection process. In addition, parameters of lines are refined so that the lines fit the data point pattern better than by using the conventional sequential iteration method.