Open Access
ARTICLE
Gender Recognition Based on Computer Vision System
Li-Hong Juanga, Ming-Ni Wub, Shin-An Linb
a School of Electrical Engineering and Automation, Xiamen University of Technology, No.600, Ligong Road, Jimei, Xiamen, 360124, p. R.China;
b Department of Information management, national taichung university of technology, taichung, taiwan RoC
* Corresponding Author: li-Hong Juang,
Intelligent Automation & Soft Computing 2018, 24(2), 249-256. https://doi.org/10.1080/10798587.2016.1272777
Abstract
Detecting human gender from complex background, illumination variations and objects under
computer vision system is very difficult but important for an adaptive information service. In this
paper, a preliminary design and some experimental results of gender recognition will be presented
from the walking movement that utilizes the gait-energy image (GEI) with denoised energy image
(DEI) pre-processing as a machine learning support vector machine (SVM) classifier to train and extract
its characteristics. The results show that the proposed method can adopt some characteristic values
and the accuracy can reach up to 100% gender recognition rate under combining the horizontal added
vertical feature and using a normal image size and test data when people are walking at a fixed angle.
Meanwhile, it will be able to achieve over 80% rate within some allowed fault-tolerant angle range.
Keywords
Cite This Article
L. Juang, M. Wu and S. Lin, "Gender recognition based on computer vision system,"
Intelligent Automation & Soft Computing, vol. 24, no.2, pp. 249–256, 2018. https://doi.org/10.1080/10798587.2016.1272777