Face recognition has become a fascinating field for researchers. The motivation behind the enormous interest in the topic is the need to improve the accuracy of many real-time applications. The complexity of the human face and the changes due to different effects make it more challenging to design as well as implement a powerful computational system for human face recognition. In this paper, we presented an enhanced approach to improve human face recognition using a back-propagation neural network (BPNN) and features extraction based on the correlation between the training images. A key contribution of this paper is the generation of a new set called the T-Dataset from the original training data set, which is used to train the BPNN. We generated the T-Dataset using the correlation between the training images without using a common technique of image density. The correlated T-Dataset provides a high distinction layer between the training images, which helps the BPNN to converge faster and achieve better accuracy. Data and features reduction are essential in the face recognition process, and researchers have recently focused on the modern neural network. Therefore, we used a local binary pattern histogram descriptor to prove that there is potential improvement even using traditional methods. We applied five distance measurement algorithms and then combined them to obtain the T-Dataset, which we fed into the BPNN. We achieved higher face recognition accuracy with less computational cost compared with the current approach by using reduced image features. We test the proposed framework on two small data sets, the YALE and AT&T data sets, as the ground truth. We achieved tremendous accuracy. Furthermore, we evaluate our method on one of the state-of-the-art benchmark data sets, Labeled Faces in the Wild (LFW), where we produce a competitive face recognition performance.

Enhanced Human Face Recognition Using LBPH Descriptor, Multi-KNN, and Back-Propagation Neural Network

OD-LBP: Orthogonal difference-local binary pattern for Face Recognition

The invention discloses an image feature extraction method under complex illumination An improved anisotropic LOG operator with a variable scale and a variable angle as a whole and directional sensitivity is used to calculate a differential excitation component, at the same time, a gray value of a pixel pair which is symmetric with respect to a central pixel, the adaptive selection of a thresholdis achieved based on the Weber criterion, finally a two-dimensional AWASTP histogram is fused and generated, the histogram is further transformed into a one-dimensional histogram, thus the KNN algorithm is used to carry out classification, the feature dimension is reduced by comparing the gray value of the pixel pair symmetric with respect to the central pixel, thus the selection of the thresholdis correspondingly changed according to the regional features of an image, features with more discriminating ability are effectively extracted, and the recognition rate of a complex illumination image is improved The method is simple, the more texture information and gradient information in multiple directions can be extracted, thus the features of the complex illumination image are described well, and the performance is better than that of other existing algorithms

Image feature extraction method under complex illumination

Face recognition has become a fascinating field for researchers. The motivation behind the enormous interest in the topic is the need to improve the accuracy of many real-time applications. The complexity of the human face and the changes due to different effects make it more challenging to design as well as implement a powerful computational system for human face recognition. In this work, we presented an enhanced approach to improve human face recognition using a Back-Propagation Neural Network (BPNN) and features extraction based on the correlation between the training images. A key contribution of this work is the generation of a new set called the T-Dataset from the original training dataset, which is used to train the BPNN. We generated the T-Dataset utilizing the correlation between the training images without using a conventional technique of image density. The correlated T-Dataset provides a high distinction layer between the training images, which helps the BPNN to converge faster and achieve better accuracy. Data and features reduction is essential in the face recognition process, and researchers have recently focused on the modern Neural Network (NN). Therefore, we used a Principal Component Analysis (PCA) descriptor to prove that there is a potential improvement even using traditional methods. We applied five distance measurement algorithms and then combined them to obtain the T-Dataset, which we fed into the BPNN. We achieved higher face recognition accuracy with less computational cost compared to the current approach by using reduced image features. We test the proposed framework on two small datasets, the YALE and AT&T datasets, as the ground truth. We achieved tremendous accuracy. Furthermore, we evaluate our method on one of the state-of-the-art benchmark datasets, Labeled Faces in the Wild (LFW), where we produce a competitive face recognition performance. Also, we proposed an enhanced framework to improve the face registration using deep transfer learning. We used pre-trained models such as ResNet50 and SeNet50 which trained on the VGGFace2 dataset. We used the LFW dataset to evaluate our method, and we achieve high accuracy.

/pdf/improving-human-face-recognition-using-deep-learning-based-pyfj1lhgdz.pdf

Improving Human Face Recognition using Deep Learning based Image Registration and Multi-Classifier Approaches

The texture feature description becomes a tremendous challenge in the field of computer vision and pattern recognition. The high-quality feature descriptor attributes some of which are unique, due to a large number of texture classes, robust against illumination variations, and low dimensional representations. A number of image feature extraction methods had been proposed, which can be divided into two categories: holistic and local image feature extraction. The holistic feature extraction method is very sensitive to changes in geometric shapes and some variations of illumination and noise. The local image feature extraction methods can effectively overcome those weaknesses. In this study, the texture features of an RGB image are built using the Local Weighting Pattern (LWP). By using the gray-level dynamic range modification technique, Fuzzy Membership Function (MF) is applied to LWP texture images to build Fuzzy-based LWP image (FLWP). From the resulting image is then used to generate a feature descriptor in the form of labels. General Terms Image Processing, Fuzzy Logic

/pdf/texture-feature-extraction-based-on-local-weighting-pattern-2a17uezck7.pdf

Texture Feature Extraction based on Local Weighting Pattern (LWP) using Fuzzy Logic Approach

Texture feature description is a remarkable challenge in the fields of computer vision and pattern recognition. Since the traditional texture feature description method, the local binary pattern (LBP), is unable to acquire more detailed direction information and always sensitive to noise, we propose a novel method based on generalized Gabor direction pattern (GGDP) and weighted discrepancy measurement model (WDMM) to overcome those defects. Firstly, a novel patch-structure direction pattern (PDP) is proposed, which can extract rich feature information and be insensitive to noise. Then, motivated by searching for a description method that can explore richer and more discriminant texture features and reducing the local Gabor feature vector’s high dimension problem, we extend PDP to form the GGDP method with multi-channel Gabor space. Furthermore, WDMM, which can effectively measure the feature distance between two images, is presented for the classification and recognition of image samples. Simulated experiments on olivetti research laboratory (ORL), Carnegie Mellon University pose, illumination, and expression (CMUPIE) and Yale B face databases under different illumination or facial expression conditions indicate that the proposed method outperforms other existing classical methods.

https://www.mdpi.com/2073-8994/8/11/109/pdf

A Novel Texture Feature Description Method Based on the Generalized Gabor Direction Pattern and Weighted Discrepancy Measurement Model

The invention discloses an image description method suitable for complex light. The image description method comprises the following steps of: step I, acquiring a to-be-described image f(x,y); step II, processing the image f(x,y) through multi-level different motivation modes in a Weber local description operator and multi-level different direction modes in the Weber local description operator to obtain multi-level different motivation modes images Xi-multi(x,y) and multi-level different direction modes images Psi-multi(x,y); step III, carrying out segmentation on value domains of the image functions Xi-multi(x,y) and Psi-multi(x,y) which are obtained in step II to obtain functions Xi'-multi(x,y) and Psi'-multi(x,y); step IV, calculating histograms of the functions Xi'-multi(x,y) and Psi'-multi(x,y) which are obtained in step III to obtain functions as shown in the specification; and step V, connecting the functions as shown in the specification to obtain a final image description as shown in the specification. According to the image description method, target image characteristics in a complex light environment can be effectively extracted, and the image description method is suitable for application in the image recognition field in the complex light environment.

Xiangmo Zhao

Papers

A Novel Texture Feature Description Method Based on the Generalized Gabor Direction Pattern and Weighted Discrepancy Measurement Model

Image description method suitable for complex light