Locating human faces within images
Introduction
Locating human faces in images generated by cameras is a complicated task to automate [21]. The reason for this is that there are several variables involved which can affect image of a face. The hindrance to face detection is the presence of additional features such as eyeglasses, moustache, or beard. The presence of these features can greatly distort the basic image of a face. In addition, the surrounding environment can also pose a problem. Changes in light-source distribution can cast or remove significant shadows from a particular face, therefore bringing more variability to facial patterns. A face locator should cope with the wide range of allowable facial pattern variations in an image.
In this paper, an intelligent face locator is proposed to be used for facial tracking in various applications including humanoid robots. The proposed face locator can deal with different variations in an image. The face locator improves the performances of the existing systems by accommodating for several variations in the input image such as illumination, expression, pose, facial-hair, occlusion, noise, eyeglasses, and background. Some of these image variations, e.g. illumination, are handled through including examples of the image variations in the training sets. Other image variations are handled through specifically design of the face locator.
The face locator consists of three modules: preprocessing, face-components extraction, and final decision-making. In the first module, standard image preprocessing is performed to enhance and restore the raw input image. Face components are then extracted in the second module using fuzzy neural networks. In the third module, final evaluation of the identified face components and determination of the face locations are performed using a commonsense-knowledge base.
This paper is organised as follows. The existing face-detection methods are reviewed in Section 2. In Section 3, the architecture of the proposed system is given in detail. The experimental results on a large collection of several test images are presented in Section 4. In Section 5, the performance of the face locator is compared against those of two well-known existing face-detection systems. In Section 6, concluding remarks are given.
Section snippets
Review of existing methods
To overcome the difficulties in accounting for a wide range of variations in face images, several face-detection methods have been proposed in the literature. Some of the exiting methods are briefly summarised below.
Architecture of proposed face locator
The proposed face locator (see Fig. 1) consists of three modules: preprocessing, face-components extraction, and final decision-making. These modules are explained in detail in the following sections.
Experimental results
Experiments are carried out on a large set of images, most of which are collected from the Internet. Using Rowley et al.’s images and the images collected by the author, the following seven test sets are constructed for evaluating the performance of the proposed face locator:
Test set 1. Contains 107 scanned photographs, newspaper pictures, images collected from the Internet, and digitised television pictures put together by Rowley et al. for testing their first system [14]. These images contain
Discussions
Table 1, Table 2 illustrates the performance of the face locator on Test sets 1–7. Included in the tables are the accuracy results for Sung–Poggio’s method on Test set 2. In addition, the tables show the performance of the Rowley et al. 1 and Rowley et al. 2 on Test sets 1–4. Sung and Poggio assume that the number of faces in Test set 2 is 149, however, Rowley et al. assume that there are 155 faces in the same test set. The author also assume that the number of faces in Test set 2 is 155.
As
Concluding remarks
A face locator was proposed to improve the performances of the existing counterparts. The face locator can handle variations due to changes in pose of up to ±45° in both up/down and left/right directions in a 3D space, rotation of up to ±45° in frontal direction in a 2D plane, illumination, expression, facial-hair, occlusion, and existence of noise and eyeglasses.
The performance of the face locator was evaluated by conducting experiments on seven large test sets. The performance of the face
Acknowledgements
The author would like to thank DEETYA and FUSA for their financial supports, and Fangpo He and Karl Sammut for their help.
References (22)
- H.M.E. Bakry, Face detection using neural networks and image decomposition, in: Proceedings of the 2002 International...
- et al.
Face recognition: features versus templates
IEEE Trans. Pattern Anal. Machine Intell.
(1993) - S. Edelman, D. Reisfeld, Y. Yeshurun, Learning to recognise faces from examples, Technical Report, Dept. of Applied...
- R. Feraud, O. Bernier, J.E. Viallet, M. Collobert, A fast and accurate face detector for indexation of face images, in:...
- et al.
Face detection in color images
IEEE Trans. Pattern Anal. Machine Intell.
(2002) - T.S. Jebara, A. Pentland, Parametrized structure from motion for 3D adaptive feedback tracking of faces, in:...
- A.Z. Kouzani, Invariant face recognition, Ph.D. Thesis, Engineering, Flinders University, Australia,...
- A.Z. Kouzani, A. Bouzerdoum, A generic fuzzy neuron, in: Proc. 1995 IEEE/Nagoya University World Wsepersons Workshop on...
- A.Z. Kouzani, F. He, K. Sammut, Commonsense knowledge representation and reasoning with fuzzy neural networks, in:...
- A.Z. Kouzani, S. Nahavandi, L.X. Kong, F.H. She, Pixel-based correspondence for facial images, in: Journal of Images...
Cited by (9)
Face recognition algorithm based on algebraic features of SVD and KL projection
2016, Proceedings - 2016 International Conference on Robots and Intelligent System, ICRIS 2016A framework for making face detection benchmark Databases
2014, IEEE Transactions on Circuits and Systems for Video TechnologyMulti-skin color clustering models for face detection
2010, Proceedings of SPIE - The International Society for Optical EngineeringSemantic processing of color images
2006, Color Image Processing: Methods and ApplicationsAn ontology-based framework for semantic image analysis and retrieval
2006, Semantic-Based Visual Information RetrievalLearning probabilistic classifiers for human-computer interaction applications
2005, Multimedia Systems