In this paper, we propose a scheme for moving object tracking from videos by combining mean shift and motion field statistics. For mean shift, we employ an enhanced spatial-range mean shift that enables a reduced numb...
详细信息
In this paper, a novel neural network based manifold learning method(NNBML)[1] recently appeared in the Journal of Science is introduced. It can effectively convert high-dimensional data into low-dimensional codes, wh...
详细信息
ISBN:
(纸本)9781601320438
In this paper, a novel neural network based manifold learning method(NNBML)[1] recently appeared in the Journal of Science is introduced. It can effectively convert high-dimensional data into low-dimensional codes, which are then used for classification. However, it performs not well while dealing with small size face database used for face recognition. We propose a solution generating more samples data based on the existing data. The proposed method is implemented on two well-known face databases, viz. ORL and Yale face databases. The experimental results show that NNBML is able to deal with the task of face recognition after more data samples generated using the proposed method, and also that NNBML outperforms LDA in terms of recognition rate.
In face recognition, the dimensionality of raw data is very high, dimension reduction (Feature Extraction) should be applied before classification. There exist several feature extraction methods, commonly used are Pri...
详细信息
In this paper, a face recognition method using local qualitative representations is proposed to solve the problem of face recognition in varying lighting. Based on the observation that the ordinal relationship between...
详细信息
ISBN:
(纸本)9780819469526
In this paper, a face recognition method using local qualitative representations is proposed to solve the problem of face recognition in varying lighting. Based on the observation that the ordinal relationship between the average brightness of image regions pair is invariant under lighting changes, Local Binary Mapping is defined as an illumination invariant for face recognition based on Local Binary pattern descriptor, which extracts the local variance features of an image. For the 'symbol' feature vector, hamming distance is used as similarity measurement. It has been proved that the proposed method can provide the accuracy of 100 percent for subset 2, 3, 4 and 98.89 percent for subset 5 of the Yale facial database B when all images in subset 1 are used as gallery.
Automated tongue image segmentation in tongue diagnosis system of traditional Chinese medicine is difficult due to two factors: There are lots of pathological details on the surface of tongue, and the shapes of tongue...
详细信息
The sliding window method will cause the severe unbalanced dataset problem. In this paper, under-sample the majority class method is adopted to solve this problem,and SVM is used to classify the processed data The bet...
详细信息
The sliding window method will cause the severe unbalanced dataset problem. In this paper, under-sample the majority class method is adopted to solve this problem,and SVM is used to classify the processed data The better prediction result of minority class (that is, the signal peptides positive sample set) is ***, we discover that the (-3,-1) rule is helpful to the prediction. So Information content based feature weighting method is proposed This method avoids the blindness of the previous algorithm in dealing with different sites. Experiments show that not only is the correct prediction rate of minority class improved dramatically, but also the correct prediction rate of majority class is kept in a high *** of the unbalanced data processing and the proposed information content based feature weighting method can greatly improve the performance of SVM classifier of signal peptides.
Linear Discriminant Analysis (LDA) is frequently used for dimension reduction and has been successfully utilized in many applications, especially face recognition. In classical LDA, however, the definition of the betw...
详细信息
In this paper, a pixel-level image fusion algorithm based on Nonsubsampled Contourlet Transform (NSCT) has been proposed. Compared with Contourlet Transform, NSCT is redundant, shift-invariant and more suitable for im...
详细信息
ISBN:
(纸本)9780819469519
In this paper, a pixel-level image fusion algorithm based on Nonsubsampled Contourlet Transform (NSCT) has been proposed. Compared with Contourlet Transform, NSCT is redundant, shift-invariant and more suitable for image fusion. Each image from different sensors could be decomposed into a low frequency image and a series of high frequency images of different directions by multi-sacle NSCT. For low and high frequency images, they are fused based on local-contrast enhancement and definition respectively. Finally, fused image is reconstructed from low and high frequency fused images. Experiment demonstrates that NSCT could preserve edge significantly and the fusion rule based on region segmentation performances well in local-contrast enhancement.
Due to ear's complex structure, particular position, and preferable stability, ear biometrics has attracted increasingly attention recently. In this paper, we present a new multi-view based ear feature extraction ...
详细信息
Due to ear's complex structure, particular position, and preferable stability, ear biometrics has attracted increasingly attention recently. In this paper, we present a new multi-view based ear feature extraction strategy. We utilize not only front view ear image but backside view ear image to extract 2D ear shape four kinds of rich features for ear recognition. In addition, we utilize multi-view ear images to reconstruct 3D ear shape, and a neural network 3D ear registration method is introduced also. Experimental results and comparison analysis show our multi-view based strategy will be a promising approach for ear biometrics.
In order to segment cells images accurately and efficiently we combine the model proposed by Tony F. Chan with Gabor filter. The model proposed by Tony F. Chan can detect very weak boundary based on function energy in...
详细信息
In order to segment cells images accurately and efficiently we combine the model proposed by Tony F. Chan with Gabor filter. The model proposed by Tony F. Chan can detect very weak boundary based on function energy instead of the gradient. Gabor filter can enhance the cells boundaries in different angle and denoise the image. In the experiments, we firstly use Gabor filter to enhance the image with different angle, and then fuse the enhanced images to get better images. At last we use the active contour model to the fused image. The result shows the model we proposed is better than the model which dose not use Gabor filter.
暂无评论