We describe a new method for recognizing humans by their face, which is robust to the variations of facial imaging conditions, with high accuracy. The human face recognition system consists of three components: i) a n...
详细信息
We describe a new method for recognizing humans by their face, which is robust to the variations of facial imaging conditions, with high accuracy. The human face recognition system consists of three components: i) a new face descriptor based on edge component histogram and its variance between pixels, ii) Gab or-edge components histogram for facial image representation, combining the Gab or wavelet and the proposed edge components histogram, iii) a sparse representation classifier for the face recognition. The effective and robust face recognition with high accuracy is achieved by the Gab or-edge components histogram and the sparse representation classifier. In experiments, higher face recognition performances, which are 99.45% on ETRI database and 99.41% on XM2VTS database, have been achieved.
In this paper we present a novel algorithm based on Super-Resolution image Reconstruction (SRIR). We use pattern recognition method to optimize the performance of digital watermarking. First, the binary watermarking i...
详细信息
The adaptive loop filter (ALF) is a Wiener algorithm basedtechnology which is applied to the reconstructed frame to improve the image quality in high-efficiency video coding (HEVC). In the conventional ALF algorithm,...
详细信息
ISBN:
(纸本)9780889869493
The adaptive loop filter (ALF) is a Wiener algorithm basedtechnology which is applied to the reconstructed frame to improve the image quality in high-efficiency video coding (HEVC). In the conventional ALF algorithm, filter coefficients are re-designed for 16 times to obtain the best sets in one frame and also to increase the external memory access and computational cost significantly. This contribution proposes an inter-channel transformation scheme based on the correlation between chroma and luma to alleviate both computing time and memory bandwidth. First, the proposed algorithm generates the cross-correlation matrix and auto-correlation matrix in chroma samples which only spends one-quarter computational costs due to the resolution difference. Then, it applies the inter-channel transformation to derive the filter coefficients. The proposed methods not only eliminate a large number of multiplication operations, but also reduce the run-time by 58.70% averagely during whole ALF encoding process, while a negligible loss 0.074 BD-PSNR in the Y component.
The proceedings contain 78 papers. The topics discussed include: topic page mining based on phraserank for advertisement image;a multi-tool scheme for summarizing textual documents;ontology on the level g of the softw...
详细信息
ISBN:
(纸本)9789898533098
The proceedings contain 78 papers. The topics discussed include: topic page mining based on phraserank for advertisement image;a multi-tool scheme for summarizing textual documents;ontology on the level g of the software process model *** to assist business processes modeling;automation of a vending system using smartphones;heuristic and AI approach to optimize plagiarism detection tool using a public search engine;an incremental approach to technology-supported curriculum design and approval;social network engineering and ontology engineering for e-learning: how do these work together?;a modeling approach for knowledge management in complex business systems;on the use of social media for important things: Facebook as a coping tool;meaningful learning in mathematics education: a proposal of developing a prototype of an augmented reality tool;distinction between opinion and information spreading in social networks;and Genapi: a generic social-networking API.
Universal steganalysis detect the presence of secret messages. In this paper, we proposed a new methodology which will be created different groups of images based on image features as an unsupervised learning and prop...
详细信息
This paper deals with the detection of skin pixels in color images containing different ethnic groups' skins. These images are subject to any form of distortion such as the variation of illumination and the variet...
详细信息
This paper deals with the detection of skin pixels in color images containing different ethnic groups' skins. These images are subject to any form of distortion such as the variation of illumination and the variety of capture devices. We have used combination scheme of linear classifiers where each one is devoted to a specific ethnic group. We used kernel Bayesian logistic regression because it outperforms many existing linear classifiers. It is shown that our scheme outperforms some existing skin detection methods that provide high classification scores.
This paper proposes a modified structure of Retinex algorithm for the image enhancement method. The proposed algorithm performs robustly and stably on various environments, particularly in bad weather and night vision...
详细信息
This paper proposes a modified structure of Retinex algorithm for the image enhancement method. The proposed algorithm performs robustly and stably on various environments, particularly in bad weather and night vision. Especially, Retinex algorithm tends to show outstanding results as compared with other approaches, and also shows advantages at night vision. However, the Retinex algorithm has problems such as the halo effect and the color distortion. In this paper, the proposed algorithm has a simple idea to remove above problems that combines a reflectance and a luminance image with some constant factors respectively. Since the result includes the luminance component, it works on diverse environment such as not only foggy image but also night vision image. Furthermore, this algorithm has proper structure for real time processing. Results on a variety of images show the effectiveness of this approach.
In order to overcome the large computing of Non- Local (NL) means for image de-noise, a new method of FANL(Fuzzy Adaptive Non Local) means de-noise is proposed in this paper. It is proposed that the noise variance is ...
详细信息
In this paper, we propose two semi-dynamic calibration methods for compensating for user's head movements for an eye gaze pointing system. Since the user perceives degradation in pointing accuracy during use, an e...
详细信息
ISBN:
(纸本)9789898565259
In this paper, we propose two semi-dynamic calibration methods for compensating for user's head movements for an eye gaze pointing system. Since the user perceives degradation in pointing accuracy during use, an effective compensatory calibration by the user which does not require additional apparatus or high cost calculation can be a useful solution for the problem. The proposed semi-dynamic calibration methods lead the user to gaze at 1 or 3 points on the computer screen and reduce the gap between the true eye gaze direction and the position of the mouse pointer. Experimental results showed that the proposed methods were capable of pointing the mouse pointer within 20 pixels at a distance of about 60 cm between the user and the display.
In these days, video technology has successfully improved creating tremendous results. As video technologies improve, multimedia devices and demands from users are diversified. Therefore, the video codec used in these...
详细信息
In these days, video technology has successfully improved creating tremendous results. As video technologies improve, multimedia devices and demands from users are diversified. Therefore, the video codec used in these devices should support various displays with different resolutions. The technology to generate a higher resolution image from the associated low-resolution image is interpolation. The interpolations are generally performed in either the spatial domain or the DCT domain. To use the advantages of both domains, we proposed advanced hybrid interpolation algorithm based on edge direction, which adaptively uses interpolation in both domains. The experimental results demonstrate that our algorithm has good performance in terms of PSNR with reducing blocking artifacts.
暂无评论