Multiscale facial structure representation for face recognition under varying illumination

Zhang, Taiping; Fang, Bin; Yuan, Yuan; Tang, Yuan Yan; Shang, Zhaowei; Li, Donghui; Lang, Fangnian

doi:10.1016/j.patcog.2008.03.017

Cited by 170 publications

(66 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the Retinex theory, a face image is decomposed into its smoothed version and its illumination invariant features. To obtain such decomposition, the total variation model, Gauss filtering, weighted Gauss filtering, and wavelet transform are employed in logarithmic total variation (LTV) model [11], intrinsic image [45], self quotient image (SQI) [12], and logarithmic wavelet transform (LWT) [13], respectively. Among the methods of extracting the illumination insensitive/invariant features, the Retinex theory-based methods always perform better than the others.…”

Section: ) Category I: Extracting Illumination Insensitive Featuresmentioning

confidence: 99%

“…The PMF algorithm produces almost the same result as the SOCP algorithm, but its computation is much faster and is almost in real-time. After getting and , and can be estimated as follows: (6) There are also other feature-extraction methods for image decomposition, for example, the self quotient image (SQI) [12] and LWT [13]. The results of face image decomposition by different methods are illustrated in Fig.…”

Section: B Face Image Decompositionmentioning

confidence: 99%

See 1 more Smart Citation

Normalization of Face Illumination Based on Large-and Small-Scale Features

Xie

Zheng

Lai

et al. 2011

IEEE Trans. on Image Process.

134

View full text Add to dashboard Cite

Abstract-A face image can be represented by a combination of large-and small-scale features. It is well-known that the variations of illumination mainly affect the large-scale features (low-frequency components), and not so much the small-scale features. Therefore, in relevant existing methods only the small-scale features are extracted as illumination-invariant features for face recognition, while the large-scale intrinsic features are always ignored. In this paper, we argue that both large-and small-scale features of a face image are important for face restoration and recognition. Moreover, we suggest that illumination normalization should be performed mainly on the large-scale features of a face image rather than on the original face image. A novel method of normalizing both the Small-and Large-scale (S&L) features of a face image is proposed. In this method, a single face image is first decomposed into large-and small-scale features. After that, illumination normalization is mainly performed on the large-scale features, and only a minor correction is made on the small-scale features. Finally, a normalized face image is generated by combining the processed large-and small-scale features. In addition, an optional visual compensation step is suggested for improving the visual quality of the normalized image. Experiments on CMU-PIE, Extended Yale B, and FRGC 2.0 face databases show that by using the proposed method significantly better recognition performance and visual results can be obtained as compared to related state-of-the-art methods.

show abstract

Section: ) Category I: Extracting Illumination Insensitive Featuresmentioning

confidence: 99%

Section: B Face Image Decompositionmentioning

confidence: 99%

Normalization of Face Illumination Based on Large-and Small-Scale Features

Xie

Zheng

Lai

et al. 2011

IEEE Trans. on Image Process.

134

View full text Add to dashboard Cite

show abstract

“…Retinex method [13][14][15][16][17][18][19][20][21] is used for image decomposition. The input face image is decomposed as R and L respectively Reflectance and Illumination by applying retinex method.…”

Section: Decompositionmentioning

confidence: 99%

“…To obtain such decomposition, the total variation model, Gauss filtering, weighted Gauss filtering, and wavelet transform are employed in logarithmic total variation (LTV) model [14], self quotient image (SQI) [15], respectively. Among the methods of extracting the illumination insensitive/invariant features, the Retinex theory-based methods [13][14][15][16][17][18][19][20][21] always perform better than the others. However, the large-scale features of a face image, which may also contain useful information for recognition, are always discarded in these methods.…”

Section: Introduction and Relevant Previous Workmentioning

confidence: 99%

Illumination Invariant Face Recognition

Kaur¹,

Kaur²

2013

IJCA

View full text Add to dashboard Cite

Face Recognition is a term that includes several subproblems. Though face recognition have been a grown up research area, however, there still remain many problems that must be overcome to develop a robust face recognition system that works well under various circumstances such as illumination, pose and expressions variations. Such variations have proven to be one of the biggest problems of face recognition systems. In the proposed thesis work the problem of illumination is discussed. A method based on the combination of Retinex and LOG-DCT technique is applied to suppress the illumination and for better face recognition results. After illumination normalization LBP is used for extracting the features of normalized images, which are further used for face recognition. In the proposed method the experiments on Extended Yale B database show that by using the proposed method better recognition performance and results can be obtained.

show abstract

“…To alleviate the adverse effect of complex illumination on face recognition, researchers proposed many methods such as multiscale retinex (MSR) [1], self quotient image (SQI) [2], logarithmic total variation (LTV) [3], wavelet-based methods [4]- [6], gradient-based methods [7]- [9], Gabor-based methods [10], Weberfaces-based methods [11], oriented local histogram equalization (OLHE) [12], the small-and large-scale (S&L) features [13] and enhanced local texture feature [14], [15]. However, most of them are still sensitive to illumination variation.…”

Section: Introductionmentioning

confidence: 99%

A Novel Illumination Estimation for Face Recognition under Complex Illumination Conditions

Cheng

Han

2017

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYAfter exploring the classic Lambertian reflectance model, we proposed an effective illumination estimation model to extract illumination invariants for face recognition under complex illumination conditions in this paper. The estimated illumination by our method not only meets the actual lighting conditions of facial images, but also conforms to the imaging principle. Experimental results on the combined Yale B database show that the proposed method can extract more robust illumination invariants, which improves face recognition rate.

show abstract

Multiscale facial structure representation for face recognition under varying illumination

Cited by 170 publications

References 24 publications

Normalization of Face Illumination Based on Large-and Small-Scale Features

Normalization of Face Illumination Based on Large-and Small-Scale Features

Illumination Invariant Face Recognition

A Novel Illumination Estimation for Face Recognition under Complex Illumination Conditions

Contact Info

Product

Resources

About