An Enhanced Approach of Face Detection using Bacteria Foraging Technique

An Enhanced Approach of Face Detection using Bacteria Foraging Technique

Kapil Kumar Gupta (Computer Science Department, Goel Institute of Technology and Management Lucknow, Lucknow, India), Rizwan Beg (Integral University, Lucknow, India) and Jitendra Kumar Niranjan (FreeKaaMaal, Noida, India)
Copyright: © 2016 |Pages: 11
DOI: 10.4018/IJCVIP.2016010101


In this study, authors present an enhanced approach of face detection using bacteria foraging technique. This approach is based on chemotexis, reproduction and elimination and dispersal step. In this study the authors analysed face detection algorithm based on human skin color and fitting the ellipse as human face can be approximate by ellipse. Their approach for face detection requires no initial pre-processing of the image. A number of Bacteria agents with evolutionary behaviours are uniformly distributed in the 2-D image environment to search the skin-like pixels and locate each face-like region by evaluating the local color distribution. This approach has the advantage of very fast face detection by reducing pre-processing time of the image. This approach significantly improves face detection rate.
Article Preview


Face Detection is a very important task within the field of computer vision. It is the first essential step in a complete face recognition system. Furthermore, it has its applications in security systems, multimedia systems and human-computer interaction systems, i.e. a computer monitor flashing on, only when a person is in front of it, and going into power saving mode when the person leaves. Human face localization and detection is often the first step in applications such as video surveillance, human computer interface, and image database management.

The process of human face detection is a complex problem to computer but if sufficient information provided, it will be able to detect human face. One of the most important features for detecting the face is the skin color of human face. However, color is not sufficient to detect the human and localize it because other body part may have similar color. Therefore, we need some other features such as geometry of the face and eyes location.

Many Face detection techniques are reported in the literature of image processing mechanisms. Face detection in gray images based on edge orientation feature have proposed by Froba et al. (2002). Another technique has been proposed by Nara et al. (2004) in color image using the shape of face and edges. Johg MinLee et al (2007) has worked on face detection suing edge orientation and geometric feature. Wu and Nevatia et al (2007) present an approach to automatically detect and track multiple, possibly partially occluded humans in a walking or standing pose from a single camera, which may be stationary or moving. Jinquinglui et al. (2010) suggest an improved canny and soble edge operator for the face detection. The Kirsch edge detection algorithm's shortage, a method of improved Kirsch human face edge detection was proposed by Yuweibo et al. (2011). Many researchers have worked on face detection using the template matching. Ben Zion et al (2003) investigate the problem of detecting human faces in complex images using large set of templates and hash table for improving the searching efficiency. Seo et al. (1999) and Wang et al. (2000) have proposed new method for the face detection using the template matching. Jianguo et al (1997) propose a new method based on shape information and edge template matching. A classic example of appearance-based methods, Mogghadam et al. (2002) used Eigen- space representation for individual features, i.e. the eyes and mouth. However, this method was also effected by rotation and illumination. Another category is the invariant feature-based approach. These methods attempt to find consistent features that are discriminative enough even when the pose, viewpoint, or illumination conditions could vary. Among many approaches involving the invariant-based method for facial features, the facial color and the edge property are two typical invariants. The facial color characteristic is a simple and fast approach for pose-variant face detection. In Tian et al. (2000) and Le et al. (2006), which use facial color information with lighting compensation, a chromatic map of the facial features is created. Taigun Lee et al. (1997) proposed new method based on pattern of facial feature in color images. Kinchoong et al. (1999) suggest new methods which adopt a bottom-up feature-based approach which has the flexibility to be extended to different scale, orientation and viewpoint of faces in the image. The neural network approach detects faces by sub-sampling different regions of the image to a standard- sized sub image and then passing it through a neural network filter. Some authors (Rowley, 1999; Fasel, 1998; Lin et al., 1997; Feraud et al. 1997) use artificial neural network trained with the normalized face image but these method requires more computation time. Generative neural network models have been proposed by Raphael Feraud et al. (1999) based on neural network. Another category of face detection is using fuzzy pattern matching and neuro-fuzzy classifier. Haiyuan et al. (2010) proposed a fuzzy theory based on pattern matching technique and use it to detect a face candidate finding out pattern similar to prebuild head shape model from the extracted skin and hair region. Akihiro yorita et al. (2007) suggests a fuzzy based using the genetic algorithm an optimization technique. Recently, evolutionary computation has been applied to improve the performance of image processing (Silistijono et al., 2007; Kim et al., 2005).

Complete Article List

Search this Journal:
Open Access Articles: Forthcoming
Volume 11: 4 Issues (2021): Forthcoming, Available for Pre-Order
Volume 10: 4 Issues (2020): 3 Released, 1 Forthcoming
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 2 Issues (2016)
Volume 5: 2 Issues (2015)
Volume 4: 2 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing