Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

A new method for occluded face detection from single viewpoint of head

Authors: Theekapun Charoenpong; Chaiwat Nuthong; Ukrit Watchareeruetai;

A new method for occluded face detection from single viewpoint of head

Abstract

Due to a problem of current research concerning with occlude face detection occurring when detecting occluded face captured from any viewpoint of head between -90 degrees to +90 degrees, we propose a new method to detect occluded face from a viewpoint of face by skin color ratio of two parts of head region. Head data is captured from any viewpoint between -90 degrees to +90 degrees of viewpoint of head. This method consists of four steps which are primary head regions extraction, head area identification, skin area segmentation, and classification. For first step, foreground is extracted by Mahalanobis distance and background subtraction. In second step, head area is extracted based on primary head region. In third step, skin area is segmented by using multi-skin color database. Head region is divided into two parts based on center of head. For fourth step, a criterion of skin ratio of two parts of head is used for classification. In this paper, occluded face is detected by a criterion of skin ratio from each side of head. To evaluate performance of the method, huskin color ratio of two parts of head region. Head data is captured from any viewpoint between -90 degrees to +90 degrees of viewpoint of head. This method consists of four steps which are primary head regions extraction, head area identification, skin area segmentation, and classification. For first step, foreground is extracted by Mahalanobis distance and background subtraction. In second step, head area is extracted based on primary head region. In third step, skin area is segmented by using multi-skin color database. Head region is divided into two parts based on center of head. For fourth step, a criterion of skin ratio of two parts of head is used for classification. In this paper, occluded face is detected by a criterion of skin ratio from each side of head. To evaluate performance of the method, human head with and without obstacle captured from any viewpoint of headman head with and without obstacle captured from any viewpoint of head between -90 degrees to +90 degrees around the head is used. Based on a criterion of skin ratio from two sides of head, accuracy rate of non-occluded face and occluded face detection is 86.29%, and 91.02%, respectively. Advantage of this method is that this method can detect occluded face such as helmet or mask from any viewpoint of head.

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    5
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
5
Average
Average
Average
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!