As a guest user you are not logged in or recognized by your IP address. You have
access to the Front Matter, Abstracts, Author Index, Subject Index and the full
text of Open Access publications.
Given a set of points, a convex hull is the smallest convex polygon containing all the points. In this paper, we describe a high-speed method for calculating the convex hull of 2D images based on Andrew's monotone chain algorithm using FPGA. This algorithm arranges input points in ascending order, according to their y-coordinates, and repeatedly checks for convexity using every group of three subsequent points by calculating the cross product of the vectors they generate. In order to arrange the points in ascending order, they must be sorted by their y-coordinates, which tends to significantly increase the total execution time when calculating the convex hull in larger images. In our method, (1) all the points on a row in a given image are acquired in parallel, (2) only the points that can compose the convex hull are selected, and (3) the convex hull is concurrently built from the selected points during the previous steps. These three steps are successively executed until all rows in the image have been processed. The FPGA implementation of our method produces throughput 23.2 times faster than the GPU implementation.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.
This website uses cookies
We use cookies to provide you with the best possible experience. They also allow us to analyze user behavior in order to constantly improve the website for you. Info about the privacy policy of IOS Press.