Highlights

  • ISSN Number : 2584-0673

  • Four Issues Per Year

  • No Publication Fee 

  • Free and Open Access Peer Reviewed Journal  

  • Articles Invited in English only 

  • Submission to a Final Decision: 45 Days

  • Acceptance to Publication: 12 Days

Track manuscript
Enter only the numeric value. (eg. Enter 2014000001, not IJIMSR2014000001) and click on Track
Abstract
Fingerprint Based Blood Group Identification Using HOG Feature Extraction with Machine Learning - Volume -3 | Issue - 1 | 2025, (OCT - DEC)

Pages:9-24

Category: Engineering & Technology

Published Date: 31-Dec-2025

Graphical Abstract

P. Srinivasa Rao, Kuppili Jithi, Kolukuluri Sowmya, K. Sobha Rani

Keywords:

Histogram of Oriented Gradients (HOG), Fingerprint Feature Extraction, Support Vector Machine (SVM), Blood Group Prediction, Random Forest, Biometric-Based Diagnosis

Abstract:
Identifying blood groups plays crucial role in medical diagnosis, blood transfusions and emergency healthcare. The usual ways to determine blood typification style it rely on serologic tests. To perform these tests, we need medical experts to collect blood samples and to testing blood sample they need reagents, and have access to a laboratory. Usually, they take time, more money and they use the patient's blood in more quantity. To address this, this study examines a new, non-invasive approach to detect blood groups using fingerprint analysis. Both fingerprints and blood groups come from our genes. So, the patterns in fingerprints like loops, whorls, and arches might link to blood group traits. The new system involves extracting fingerprint features using a Histogram of oriented gradients (HOG), which is a widely used feature descriptor for pattern recognition. These extracted features are then classified using the support vector machine (SVM) and random forest (RF) to predict blood groups. The study evaluates the performance of the model using accuracy, precision, recall, F1 score and confusion matrices. The HOG-SVM achieves high accuracy of 90.69% and HOG-RF model achieves accuracy of 87.36%.  In contrast to prior statistical or manual correlation studies, our method applies a fully-automatic machine learning-based pipeline, obtains an accuracy of 90% irrespective of blood samples and laboratory testing. The study results that HOG-SVM performs high classification accuracy compared to HOG-Random Forest. 
Announcements

Call for Paper

Next Issue

Last Date : 30th April 2026

Search Articles

Indexed and Abstracted in

11_images.png
13_o7.png
14_o4.png
16_Open_Access_logo_with_dark_text_for_contrast,_on_transparent_background.png
17_ORCID_full_logo.png