Open AccessReal Time Detection of Object Blob Localization Application using 1-D Connected Pixel and Windowing Method on FPGA
Author(s) -
Chee Kiang Lam,
Phaklen Ehkan,
Rafikha Aliana A. Raof,
Suwimol Jungjit
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1755/1/012054
Subject(s) - field programmable gate array , pixel , computer science , artificial intelligence , computer vision , process (computing) , kernel (algebra) , blob detection , object detection , image (mathematics) , computer hardware , edge detection , image processing , pattern recognition (psychology) , mathematics , combinatorics , operating system
Blob detection and localization is a common process used in the machine vision. Current existing blob detection method is using 2-dimensional kernel matrix which is higher in time consumption and also memory space. This study has proposed a dedicated digital architecture consist of two modules to detect binary image blob using only 1 -dimensional image pixel. First module is used to detect connected pixel in a row of pixel, and second module is used to perform windowing to justify blob location. This study has been successfully implemented and tested on Altera DE2 FPGA board. The proposed architecture only takes 24 clock cycles to deliver blob location and related features. The tested architecture only utilizes 1597 logic element, or 4.81% of the FPGA total resources.