Title :
A region merging approach for image segmentation on FPGA
Author :
Trieu, Dang Ba Khac ; Maruyama, Tsutomu
Author_Institution :
Syst. & Inf. Eng., Univ. of Tsukuba, Tsukuba, Japan
Abstract :
Image segmentation is one of the most important tasks in the image processing, and many algorithms for the segmentation have been proposed. In the segmentation based on the mean-shift, k-means and so on, the image is once over-segmented, and then the small regions are merged. In this paper, we propose a region merging algorithm for hardware systems, in which the data are not managed strictly, and the redundant computation caused by this loose management is hidden by the pipelined and parallel processing. We have implemented the algorithm on Xilinx XC4VLX160, and its performance is about 80 fps for 768 × 512 pixel images.
Keywords :
field programmable gate arrays; image segmentation; parallel processing; FPGA; Xilinx XC4VLX160; hardware systems; image processing; image segmentation; k-means algorithm; mean-shift; parallel processing; pipelined processing; pixel images; region merging approach; Field programmable gate arrays; Hardware; Image color analysis; Image segmentation; Indexes; Merging; Real-time systems;
Conference_Titel :
Field Programmable Logic and Applications (FPL), 2012 22nd International Conference on
Conference_Location :
Oslo
Print_ISBN :
978-1-4673-2257-7
Electronic_ISBN :
978-1-4673-2255-3
DOI :
10.1109/FPL.2012.6339166
