Framework for Automatic Blood Group Identification and Notification Alert System

Boby Dewal; Madeha Memon; Mahaveer Rathi; Yasra Memon; Knooz Fatima

doi:10.33317/ssurj.578

Author(s)

Boby Dewal Department of Computer Systems Engineering, Mehran University of Engineering and Technology,Jamshoro,PakistanRaspberry Pi,Jamshoro,Pakistan

Madeha Memon Department of Computer Systems Engineering, Mehran University of Engineering and Technology,Jamshoro,PakistanRaspberry Pi,Jamshoro,Pakistan

Mahaveer Rathi Department of Computer Systems Engineering, Mehran University of Engineering and Technology,Jamshoro,PakistanRaspberry Pi,Jamshoro,Pakistan

Yasra Memon Liaquat University of Medical and Health Sciences,Jamshoro,PakistanRaspberry Pi,Jamshoro,Pakistan

Knooz Fatima Department of Computer Systems Engineering, Mehran University of Engineering and Technology,Jamshoro,PakistanRaspberry Pi,Jamshoro,Pakistan

Abstract

Image Processing has assisted researchers in a variety of ways, especially in the areas of security and medicine fields. Identifying blood types in emergencies or far-off places and regions where experts have not been available is a present-day challenge. Therefore, we have developed an automatic system that will detect the blood group and notify an alert system using GSM and various image processing methods. Prior to any treatment or operation, it is necessary to determine the blood type for a transfusion of blood, even in an emergency. Currently, technicians manually conduct these tests, which can cause human mistakes. Different systems have been created to automate these tests; none have been successful in completing the analysis in time for emergencies. This project intends to create an automated system to do these tests quickly, adjusting to urgent circumstances. Initially, the slide test is performed to collect the blood images. Furthermore, various image processing methods have been performed for processing images using the PI camera. Subsequently, an alert with the patient's blood group is then generated and sent to the concerned patient or the hospital to immediately consult the patient. Unit testing and load testing were performed on 950 images at a time which yielded 97% accuracy.

Publication Details

Page(s) 84-88

DOI 10.33317/ssurj.578

Published Journal: Sir Syed University Research Journal of Engineering and Technology, Volume: 13, Issue: 2, Year: 2023

Request the full text from PASTIC

Keywords

Image Processing, Global Positioning System, gray level Co-occurrence matrix, Blood Group Identification, Raspberry Pi.