Extensive use of surveillance cameras for human tracking and observation have been fostering the research on face recognition technique for individual identification in an unconstrained environment. However, face reco...
详细信息
Extensive use of surveillance cameras for human tracking and observation have been fostering the research on face recognition technique for individual identification in an unconstrained environment. However, face recognition is a challenging task in an unconstrained environment, where the captured images are affected by illumination effect, varying poses, noise and occlusion. The main objective of this research is to improve the accuracy and processing time in extracting facial features by using the fusion of deep learning and handcrafted architecture for recognizing individuals in unconstrained conditions, thereby providing accurate information about the individuals to security systems. The proposed system consists of multi-block local binary pattern (MB-LBP) modules for extracting the handcrafted features and Convolutional Neural Network (CNN) for extracting the high-level distinctive features. The features from both modules are fused and passed through fully connected layer with Softmax classifier to identify individuals. The results show that the enhanced algorithm based on Softmax loss function aided classifier with regularization improves the accuracy and processing time for face recognition. The proposed model improves accuracy by 94.37% against 90.01% for the state-of-the-art solution. In addition to that, it improves the processing time of 307 ms against 357 ms. The proposed system focuses on fusing hand-crafted and deep learned features to extract face features accurately and thus improving the accuracy and overall performance of the proposed system in an unconstrained environment.
Face gender recognition is an important research field in computer vision and pattern recognition. Face gender recognition is the use of computer technology to analyze face images and extract effective face features, ...
详细信息
Face gender recognition is an important research field in computer vision and pattern recognition. Face gender recognition is the use of computer technology to analyze face images and extract effective face features, so as to realize the recognition of gender attributes of observation objects. In this paper, we used multi-block local binary pattern to extract gender features, and used different support vector machine learning models to process and analyze the results. The experiments show that the combination of MB-LBP algorithm and Linear SVM is better than the combination of MB-LBP algorithm and and RBF. The experimental results show that the recognition capability of MB-LBP+SVM based on the FERET database is higher than that of the SVM, KNN+SVM and PCA+LDA+ SVM.
In this paper a moving vehicle detection algorithm based on visual processing mechanism with multiple pathways is proposed, in which the multiple pathways visual processing mechanism is inspired by the biological visu...
详细信息
ISBN:
(纸本)9783319220536;9783319220529
In this paper a moving vehicle detection algorithm based on visual processing mechanism with multiple pathways is proposed, in which the multiple pathways visual processing mechanism is inspired by the biological visual system. According to the different moving directions of front vehicles, orientation selectivity of visual cortex cells is used to construct a visual processing model with three pathways. In each pathway, an AdaBoost cascade classifier is trained using a set of special samples for detection of moving vehicles. The AdaBoost cascade classifier is response to multi-block local binary patterns (MB-LBP) of vehicles. The experimental results show that the multiple pathways visual processing mechanism, compared with the single pathway AdaBoost cascade classifier and the conventional method, not only can reduce the complexity of the classifier and training time, but also can improve the recognition rate of moving vehicle.
This paper presents a pedestrian detection approach based on multi-block local binary pattern (MB-LBP) features and Weighted Region Covariance Matrix (WRCM). multistage classifiers are used to increase the processing ...
详细信息
ISBN:
(纸本)9781467356978;9781467356992
This paper presents a pedestrian detection approach based on multi-block local binary pattern (MB-LBP) features and Weighted Region Covariance Matrix (WRCM). multistage classifiers are used to increase the processing speed and reliability of the detection system. Using the modified 3-D B-spline Wavelet-Based local Standard Deviation (BWLSD) techniques, the region of interest is determined. Once the pedestrian region is identified, the front end of the multistage classifier quickly determines wherever pedestrians may be present, while the back end conforms whether the first descriptor did classify correctly. The experimental results demonstrated that our approach performed well in real-time application.
暂无评论