Open AccessHardware implementation of Sobel edge detection system for blood cells images-based field programmable gate array
Author(s) -
Anan Younis,
Basma MohammedKamal Younis,
Mohammed Sabah Jarjees
Publication year - 2022
Publication title -
indonesian journal of electrical engineering and computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.241
H-Index - 17
eISSN - 2502-4760
pISSN - 2502-4752
DOI - 10.11591/ijeecs.v26.i1.pp86-95
Subject(s) - sobel operator , field programmable gate array , computer science , edge detection , matlab , image processing , computer vision , filter (signal processing) , segmentation , enhanced data rates for gsm evolution , process (computing) , artificial intelligence , computer hardware , gate array , image segmentation , chip , field (mathematics) , image (mathematics) , mathematics , telecommunications , pure mathematics , operating system
The microscopic-blood image has been used to diagnose various diseases according to the morphological specifications of red and white blood cells. However, the manual analysis and procedures are not accurate due to the human error. Therefore, several studies conducted to find new techniques to perform this analysis using computer algorithms. The complexity of these algorithms led to thinking in simpler ways or to the hardware solutions. On the other hand, edge detection is a mathematical procedure that play an essential role in the field of medical image processing. It is considered as one of the foundations' processes for other procedures, such as the segmentation and the classification of the image. The Sobel filter is one of the conventional methods that is used to perform the edge detection process. It is based on finding the local contrast for the level of intensity of the image. This paper presents a proposed and a new method for detecting the edges of cells in the microscopic blood images using Sobel filter and its hardware implementation on the field programmable gate array (FPGA) chip. Three different techniques are proposed: MATLAB, OpenCV standard code, and FPGA customize code which give the best visual results, minimum timing results than the others.