access icon openaccess Gentle Adaboost algorithm based on multi-feature fusion for face detection

There are few types of Haar-like rectangle features, which leads to the problem that the classifier training time is too long due to the large number of feature quantities required in the description of the face. Local binary patterns (LBP) are used to describe the local texture features of the face image. Considering the inadequacy of basic LBP features in face detection, unified MB-LBP features and unified rotation-invariant LBP features are used to describe local texture features of faces. Considering the shortcomings of MB-LBP feature and rotation-invariant LBP feature on face edge information, the edge azimuth field feature based on Canny operator is combined with the above two features to describe face information. Finally, the Gentle Adaboost classifier was designed to classify all the extracted features. The experimental results show that the unified MB-LBP feature and the unified rotation invariant LBP feature and the edge azimuth field feature based on Canny operator can not only describe the face information from the local but also the whole, which greatly improves the detection rate and detection speed of the face with multiple poses and different rotation modes.

Inspec keywords: edge detection; object detection; face recognition; image classification; learning (artificial intelligence); image fusion; feature extraction; image texture

Other keywords: face image; Gentle Adaboost algorithm; face information; Canny operator; unified MB-LBP feature; local binary patterns; edge azimuth field; face edge information; face detection; multifeature fusion; local texture features; Haar-like rectangle features; Gentle Adaboost classifier; unified rotation invariant LBP feature

Subjects: Image recognition; Knowledge engineering techniques; Sensor fusion; Computer vision and image processing techniques

References

    1. 1)
      • 4. Fang, C., Su, T.: ‘Floatboost algorithm for face detection based on improved LBP features’, J. Nanjing Univ. Posts Telecommun. (Natural Science), 2014, 34, (6), pp. 7679.
    2. 2)
      • 8. Zhang, L., Chu, R., Xiang, S., et al: ‘Face detection based on multi-block LBP representation’, Adv. Biometrics, 2007, 4642, pp. 1118.
    3. 3)
      • 2. Ye, J., Zhang, Z.: ‘Face detection based on DS-adaboost algorithm’, Comput. Sci., 2013, 40, (11A), pp. 318319.
    4. 4)
      • 6. Wang, X., Zhang, Y., Chen, D.: ‘Face detection based on MB-LBP and eye tracking’, Chin. J. Sci. Instrum., 2014, 35, (12), pp. 27392744.
    5. 5)
      • 12. Ge, K.: ‘Adaboost Algorithm Based on Improved LBP Features with Skin Color Segmentation for Face Detection’, Master's thesis, Chongqing University, 2012.
    6. 6)
      • 3. Lienhart, R., Kuranov, A., Pisarevsky, V.: ‘Empirical analysis of detection cascades of boosted classifiers for rapid object detection’, Dagm, 2003, 2781, pp. 297304.
    7. 7)
      • 7. Tao, X., Wan, D., Yang, Z.: ‘A face detection method based on improved LBP feature’, Inf. Technol., 2018, 1, pp. 14.
    8. 8)
      • 9. Li, G., Li, W.: ‘Face detection under rotation in image plane using principal direction rotation LBP’, Acta Electron. Sin., 2015, 43, (1), pp. 198201.
    9. 9)
      • 5. Luan, F., Guo, S., Song, X.: ‘Adaboost fast training algorithm based on multi-feature fusion for face detection’, J. Chi. Comput. Syst., 2015, 36, (7), pp. 16131616.
    10. 10)
      • 11. Ingle, M.A., Talmale, G.R.: ‘Respiratory mask selection and leakage detection system based on canny edge detection operator’, Procedia Comput. Sci., 2016, 64, pp. 323329.
    11. 11)
      • 10. Chen, H.: ‘Multi-pose Face Detection and Recognition on Nonlinear Manifold’. Doctoral thesis, Zhejiang University, 2006.
    12. 12)
      • 1. Zhou, Z.: ‘Fixed-point optimization algorithm of AdaBoost face detection’, J. Univ. Electron. Sci. Technol. China, 2015, 44, (4), pp. 589590.
http://iet.metastore.ingenta.com/content/journals/10.1049/joe.2018.9391
Loading

Related content

content/journals/10.1049/joe.2018.9391
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading