Weber Local Descriptor (WLD) is a stable and effective feature extraction algorithm, which is based on Weber's Law. It calculates the differential excitation information and direction information, and then integrates them to get the feature information of the image. However, WLD only considers the c...
Weber Local Descriptor (WLD) is a stable and effective feature extraction algorithm, which is based on Weber's Law. It calculates the differential excitation information and direction information, and then integrates them to get the feature information of the image. However, WLD only considers the center pixel and its contrast with its surrounding pixels when calculating the differential excitation information. As a result, the illumination variation is relatively sensitive, and the selection of the neighbor area is rather small. This may make the whole information is divided into small pieces, thus, it is difficult to be recognized. In order to overcome this problem, this paper proposes Weber Symmetrical Local Graph Structure (WSLGS), which constructs the graph structure based on the $5{\times}5$ neighborhood. Then the information obtained is regarded as the differential excitation information. Finally, we demonstrate the effectiveness of our proposed method on the database of ORL, JAFFE and our own built database, high-definition infrared faces. The experimental results show that WSLGS provides higher recognition rate and shorter image processing time compared with traditional algorithms.
Weber Local Descriptor (WLD) is a stable and effective feature extraction algorithm, which is based on Weber's Law. It calculates the differential excitation information and direction information, and then integrates them to get the feature information of the image. However, WLD only considers the center pixel and its contrast with its surrounding pixels when calculating the differential excitation information. As a result, the illumination variation is relatively sensitive, and the selection of the neighbor area is rather small. This may make the whole information is divided into small pieces, thus, it is difficult to be recognized. In order to overcome this problem, this paper proposes Weber Symmetrical Local Graph Structure (WSLGS), which constructs the graph structure based on the $5{\times}5$ neighborhood. Then the information obtained is regarded as the differential excitation information. Finally, we demonstrate the effectiveness of our proposed method on the database of ORL, JAFFE and our own built database, high-definition infrared faces. The experimental results show that WSLGS provides higher recognition rate and shorter image processing time compared with traditional algorithms.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
문제 정의
Face recognition technology can be applied in a wide range of fields, such as identity authentication, access control and so on. The goal is to identify faces in a still image or video regardless of variations such as pose, illumination, occlusion and expression.
가설 설정
Specific examples are shown in Fig. 3. Our own built HD infrared face database contains 80 volunteers with different ages and gender totally. Each person has 9 photos, a total of 720 images.
제안 방법
In addition to the comparison of the recognition rate, we also test the processing time of the same image with the methods of LBP, LGS, WLD, SLGS and proposed method WSLGS. The experimental results are shown in Table 4.
In order to verify the robustness of our proposed algorithm, we also completed the experiment on JAFFE, which was devoted to facial research. The training samples for the LBP, LGS, WLD, SLGS and WSLGS algorithms are selected as 3, 4, 5, 6, 7, 8, and the experimental results are shown in Table 3.
대상 데이터
Our own built HD infrared face database contains 80 volunteers with different ages and gender totally. Each person has 9 photos, a total of 720 images. These images are collected from different angles.
In the original database, the number of images per person is different, a total of 216 images. For the convenience of testing, we selected 20 images per person, a total of 200 images for the experiment. Specific examples are shown in Fig.
ORL face database was created by the AT&T Laboratory of University of Cambridge. It contains 40 people, each person is with 10 images, there are 400 face images in total. These images include changes in posture and expression.
JAFFE database is an expression database created by Japan's ATR (Advanced Telecommunication Research Institute International). It is devoted to the study of facial expression recognition, including 10 individuals, each with 7 kinds of expressions. In the original database, the number of images per person is different, a total of 216 images.
Our experiments are performed on three databases, i.e., the ORL face database [16], our lab's high definition infrared face database and JAFFE face expression database [17].
이론/모형
On the other hand, the LGS algorithm transforms the characteristic values of the central pixels from binary to decimal format in the process of calculation, which leads to the loss of information. In order to solve these problems, we propose Weber Symmetrical Local Graph Structure algorithm, which combines these two algorithms. More specifically, WSLGS takes the 5×5 neighborhood, constructs the graph structure on two diagonal directions of the center pixel, and calculates the sum of the difference of adjacent pixels in a certain order.
In this paper, we proposed Weber Symmetrical Local Graph Structure, which is used to extract the features of face images. This algorithm constructs the graph structure from the diagonal direction in the neighborhood of 5 × 5, and then combines it with the direction with the differential excitation information according to Weber's law.
The algorithm is based on the neighborhood of 5×5. It constructs symmetric graph structure in two diagonals directions based on the original LGS algorithm and SLGS algorithm, and see it as the differential excitation in the WSLGS algorithm. Then, the differential excitation information is integrated with the information of the WLD algorithm to get the features of the center pixel.
This paper proposes the algorithm of Weber Symmetrical Local Graph Structure (WSLGS), which is inspired by WLD, LGS and SLGS. The algorithm is based on the neighborhood of 5×5.
성능/효과
67% respectively. And the recognition rate of our proposed method can reach 99.92%.
When the number of training samples is 8, the recognition rate is up to 100%. At the same time, the recognition rates of the other three algorithms are 54.37%, 57.50%, 98.63% and 59.13%.
We implemented the algorithm in the ORL database, our own built high definition infrared face database and JAFFE database. The experimental results demonstrated that the recognition rate of WSLGS algorithm is better than the traditional algorithms, and the processing time is lower than the traditional algorithms. In the future, we will try to further improve the WSLGS algorithm, hoping to be applied to other areas of biometrics, such as finger vein recognition.
It is clear from the table that the recognition rate of WSLGS algorithm is higher than the traditional algorithms. When the number of training samples is 8, the recognition rate of LBP, LGS, WLD and SLGS is 92.08%, 84.25%, 91.75% and 88.67% respectively. And the recognition rate of our proposed method can reach 99.
참고문헌 (17)
W. Zhao, R. Chellappa, P. J. Phillips, and A. Rosenfeld, "Face recognition: A literature survey," ACM computing surveys (CSUR), vol. 35, no. 4, pp. 399-458, 2003.
H. Yu and J. Yang, "A direct LDA algorithm for high-dimensional data-with application to face recognition," Pattern recognition, vol. 34, no. 10, pp. 2067-2070, 2001.
Ojala, Timo, Matti Pietikainen, and David Harwood. "Performance evaluation of texture measures with classification based on Kullback discrimination of distributions," in Proc. of Pattern Recognition, 1994. Vol. 1-Conference A: Computer Vision & Image Processing., Proceedings of the 12th IAPR International Conference on. Vol. 1. IEEE, vol. 1, pp. 582-585, 1994.
M. Heikkila, M. Pietikainen, and C. Schmid, "Description of interest regions with local binary patterns," Pattern recognition, vol. 42, no. 3, pp. 425-436, 2009.
J. Chen et al., "WLD: A robust local image descriptor," IEEE transactions on pattern analysis and machine intelligence, vol. 32, no. 9, pp. 1705-1720, 2010.
Abusham, Eimad EA, and Housam K. Bashir. "Face recognition using local graph structure (LGS)," in Proc. of International Conference on Human-Computer Interaction. Springer, Berlin, Heidelberg, pp. 169-175, 2011.
M. F. A. Abdullah, M. S. Sayeed, K. S. Muthu, H. K. Bashier, A. Azman, and S. Z. Ibrahim, "Face recognition with symmetric local graph structure (slgs)," Expert Systems with Applications, vol. 41, no. 14, pp. 6131-6137, 2014.
Y. Luo, T. Zhang, and Y. Zhang, "A novel fusion method of PCA and LDP for facial expression feature extraction," Optik-International Journal for Light and Electron Optics, vol. 127, no. 2, pp. 718-721, 2016.
Chen, Jie, et al. "A robust descriptor based on weber's law," in Proc. of Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on. IEEE, pp. 1-7, 2008.
S. Li, D. Gong, and Y. Yuan, "Face recognition using Weber local descriptors," Neurocomputing, vol. 122, pp. 272-283, 2013.
B. Wang, W. Li, W. Yang, and Q. Liao, "Illumination normalization based on weber's law with application to face recognition," IEEE Signal Processing Letters, vol. 18, no. 8, pp. 462-465, 2011.
F. S. Samaria and A. C. Harter, "Parameterisation of a stochastic model for human face identification," in Proc. of Proceedings of the Second IEEE Workshop on, presented at the Applications of Computer Vision, 1994., pp. 138-142, 1994.
M. J. Lyons, J. Budynek, and S. Akamatsu, "Automatic classification of single facial images," IEEE transactions on pattern analysis and machine intelligence, vol. 21, no. 12, pp. 1357-1362, 1999.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.