Abstract
In this paper, computational acceleration of color image segmentation using fuzzy c-means (FCM) algorithm has been presented. The color image is first converted from the Red Green Blue (RGB) color space to the YUV color space. Then, the luma (Y) information values are grouped according to the desired number of clusters using the FCM algorithm. The FCM algorithm is implemented on a Graphical Processing Unit (GPU) using the Compute Unified Device Library (CUDA) library which is developed by NVidia to speed up the computing time. Images used in this research are red blood cell images, geometry images and leaf images. The results of segmented images processed using GPU were seen identic to the results of segmented images processed using the Central Processing Unit (CPU). The computational time of the FCM algorithm can be accelerated by speed-up to 5,628 times faster and the average speed-up of all simulations done is 5,517 times faster.
Highlights
Segmentasi citra berdasarkan pada dua dasar, yaitu ketidaksinambungan dan kesamaan antar pixel
The color image is first converted from the Red Green Blue
information values are grouped according to the desired number of clusters using the Fuzzy C-Means (FCM) algorithm
Summary
Citra pada nomor 1 dan 2 merupakan citra sel darah merah, dimana sel darah merah pada citra nomor 1 dapat diamati berbentuk bulat penuh. Sedangkan pada citra nomor 2 sel darah merah terdapat beberapa sel yang dibagian tengahnya berwarna menyerupai latar belakang citra tersebut, sehingga dapat kita amati perbedaan hasil segmentasi antara citra nomor 1 dan nomor 2. Dan pada citra nomor 3, kita temui citra uji berupa bentuk geometri lingkaran berwarna merah, bujursangkar berwarna biru dan latar belakang berwarna hijau. Untuk citra nomor 4 dan 5 merupakan citra daun dengan bentuk yang berbeda. Maka hasil dari proses clustering menggunakan FCM pada citra uji dengan komputasi berbasis GPU dapat dilihat pada Tabel 2 dibawah. Hasil Simulasi Proses Clustering menggunakan FCM dengan Komputasi Berbasis GPU
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.