Vol.:(0123456789) 1 3
Journal of Ambient Intelligence and Humanized Computing
https://doi.org/10.1007/s12652-020-01886-3
ORIGINAL RESEARCH
Local Directional Maximum Edge Patterns for facial expression
recognition
V. Uma Maheswari
1
· G. Varaprasad
2
· S. Viswanadha Raju
3
Received: 14 November 2019 / Accepted: 12 March 2020
© Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract
Cognitive science and neuroscience use human facial expressions of emotion. Every single facial expression can be seen
at diferent passions in a face space. Nowadays, facial expression recognition and analysis is vital due to the demand of
introducing advanced biometric applications in every domain space. The imperative task in facial expressions of emotion
classifcation is precise feature extraction, which helps to get detailed description of facial marks. Existing feature descriptors
are sufering from various problems such as intensity variations, discrimination, vulnerability etc. In this paper, propose a
new feature descriptor method called LDMEP (Local Directional Maximum Edge Patterns) for facial expression analysis to
overcome the hindrance. We calculated the gradients in four directions of reference pixel to elicit the more feature for better
recognition instead of calculating the local diferences among neighboring pixels. We also access the orientations of the
pixels then thresholded based on the dynamic threshold to avoid the featureless area calculation. Furthermore, we considered
only dominant magnitude and orientation directions instead of all eight directions to generate feature. Thus, imperative and
efcient features are covered in dominant positions to detect the strong edges. The paper confers that how the subsequent
model can be used for the recognition of facial expression of emotion.
Keywords LDMEP (Local Directional Maximum Edge Patterns) · Facial expression analysis · Feature extraction ·
Magnitude · Phase
1 Introduction
Facial expression recognition conveys the human intention
without verbal communication. It plays a vital part in numer-
ous applications such as surveillance, medical area, psychol-
ogy, crime analysis, mood prediction, sentiment analysis
and human computer interaction. Face has the intuitive
characteristics with various expressions can be recognized
by the feature which is existed in the digital image. Thus,
an efcient feature descriptor generation from the image is
the prominent and challenging task to recognize in any cir-
cumstances, however it is tedious to retrieve exact one in all
situations. The factors depend on various parameters like
image has taken in diferent poses, illumination, background,
in various angles etc. Kotsia and Pitas (2007) facial expres-
sion recognition can be done using patterns like LBP, LTP,
DBC (Directional Binary Code) as local descriptors based
on two factors one is local features such as color, texture,
and shape etc. second is geometric based features, in con-
tent based patterns feature is extracted from the image itself
by fnding the relation between the pixel neighbors using
derivatives so much mathematical calculation is not required
to extract feature. In local features, texture based patterns
are the silver lining step in facial expressions recognition,
face recognition by extracting the feature available like fur-
rows at eyes region, beside of nose, mouth and cheeks region
on the face. In texture analysis Lucey et al. (2010) LBP is
proved and notorious to recognize face and expressions due
* V. Uma Maheswari
umamaheshwariv999@gmail.com;
umamaheswari@vardhaman.org
G. Varaprasad
varaprasad.cse@bmsce.ac.in
S. Viswanadha Raju
svraju.jntu@gmail.com
1
Department of Computer Science and Engineering,
Vardhaman College of Engineering, Hyderabad, India
2
Department of Computer Science and Engineering,
B.M.S. College of Engineering, Bangaluru, India
3
Department of Computer Science and Engineering,
JNTUHCEJ, Jagityal, Karimnagar, India